Choose Report Type
Publication Date
Report Upload
download
(2.35 MB)
vertical
Frontier Tech Hub
PDF Text
AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 1 2 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth Acknowledgements
W
e are indebted to the Expert Council for its strategic
foresight and instrumental contributions in shaping this
project. Their guidance has ensured that the roadmap
reflects both ambition and pragmatism, making the recommendations
highly actionable.
We also gratefully acknowledge the invaluable inputs of the
following domain experts: Dr Bala Subrahmanya, General Manager,
ABLE; Mr GS Krishnan, President, ABLE; Ms Kiran Mazumdar-Shaw,
Executive Chairperson, Biocon Ltd; Mr Mandar S Ghatnekar, Global
Head of IT & Digital Transformation, Biocon Biologics; Mr Shailendra
Trivedi Chief General Manager-in-Charge Department of Information
Technology, RBI and Mr Suvendu Pati Chief General Manager
Financial Technology Department, RBI; whose insights enriched the
work and firmly anchored it in ground realities.
A special word of thanks is extended to McKinsey & Company for its
exemplary partnership in providing analysis & insights that helped
us to develop a roadmap that is implementable, impact-driven, and
aligned with India’s long-term vision.
Together, the collective wisdom and collaboration of all partners
have made this effort possible. 4 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth Mr. N Chandrasekaran
Chairman, Tata Sons
Mr. Neelkanth Mishra
Chief Economist, Axis Bank and Head
of Global Research, Axis Capital
Mr. Rahul Matthan
Partner, Trilegal
Dr. Anish Shah
Managing Director and CEO,
Mahindra & Mahindra Ltd
Ms. Anu Madgavkar
Partner, McKinsey Global
Institute
Mr. Noshir Kaka
Senior Partner, McKinsey &
Company
Mr. Ishtiyaque Ahmed
Program Director, Industry/
MSME, NITI Aayog
Mr. Chandrajit Banerjee
Director General, CII
Mr. Mukesh Bansal
Founder - Myntra &
CureFit
Dr. Chintan Vaishnav
Former Mission Director, Atal
Innovation Mission, NITI Aayog
Expert Council Members I
f India is to accelerate its growth to the 8% annual rate
required for the realization of Viksit Bharat, we have
no option but to significantly raise productivity across
the economy and unlock new growth through innovation.
Artificial Intelligence can be the decisive lever. This report
sets out a practical roadmap on how we can harness Al to
translate this potential into outcomes.
The analysis highlights two major Al unlocks. First,
accelerating adoption of Al across industries to enhance
productivity and efficiency-bridging nearly 30-35% of the
required step-up. Second, transforming R&D, especially
through generative Al, which can enable India to leapfrog
into innovation-driven global opportunities, contributing
at least 20-30% of the required uplift.
With a focused and sector-specific approach, industries
such as banking and manufacturing can deploy Al today to
improve efficiency, service quality, and competitiveness-
creating momentum for deeper transformation. At the
same time, India must nurture frontier innovation, from
Al-enabled drug discovery to software-defined vehicles,
building the next engines of growth.
The path to 8% growth runs through decisive Al adoption
and innovation. This report offers a roadmap to guide that
journey. I invite government, industry, and academia to
move forward with urgency and collective purpose.
BVR Subrahmanyam
CEO, NITI Aayog
Foreword T
he NITI Frontier Tech Hub’s AI roadmap for Viksit
Bharat sends an unequivocal signal: India’s mission
to sustained 8%+ growth is anchored in bold,
pervasive AI integration and tireless innovation—and
must become a core national priority. This transformation
journey leverages sector-focused strategies and frontier
technology ecosystems, positioning India to lead the
global race in inclusive, responsible AI deployment and
governance.
Our deep gratitude is due to the Expert Council for its
strategic foresight and instrumental contributions, and
to McKinsey for its exemplary partnership in shaping a
roadmap that is implementable and impact-driven.
The time for India to lead the AI revolution at scale is now.
With robust policy frameworks, advanced infrastructure,
and collaborative innovation, India can pioneer a new
model of growth and societal advancement, ensuring
prosperity, resilience, and technological leadership for
decades to come.
The NITI Frontier Tech Hub will continue to activate this
agenda, galvanizing experts, states, and industry toward
shared progress—securing the foundations for an AI-
powered Viksit Bharat
Debjani Ghosh
Distinguished Fellow, NITI Aayog;
Chief Architect, NITI Frontier Tech Hub
Foreword Index
CHAPTER 1: INTRODUCTION....................................................................................10
AI OPPORTUNITIES FOR INDIA..........................................................................................10
STRATEGIC ENABLERS FOR AI-LED VALUE CREATION.........................................11
CHAPTER 2: POTENTIAL OUTCOMES FOR AI-LED VALUE CREATION............12
POTENTIAL OUTCOME 1: INDIA BECOMES
THE DATA CAPITAL OF THE WORLD..............................................................................13
POTENTIAL OUTCOME 2: INDIA SUPPORTS THE DEVELOPMENT OF AN
ADAPTABLE AND EFFICIENT AI-SKILLING ECOSYSTEM.......................................14
POTENTIAL OUTCOME 3: TARGETED AI ADOPTION UNLOCKS SECTORAL
GROWTH......................................................................................................................................14
POTENTIAL OUTCOME 4: INDIA’S JOBS ARE FUTURE-PROOFED, AND
INDUSTRY TRANSFORMED AT SCALE...........................................................................15
CHAPTER 3: LEVER 1 - ACCELERATING AI ADOPTION
ACROSS INDUSTRIES TO IMPROVE PRODUCTIVITY AND EFFICIENCY..........17
BANKING.....................................................................................................................................19
POTENTIAL ENABLERS FOR CONSIDERATION.....................................................21
KEY RISKS..............................................................................................................................23
MANUFACTURING....................................................................................................................24
POTENTIAL ENABLERS FOR CONSIDERATION.....................................................25
KEY RISKS..............................................................................................................................26 CHAPTER 4 – LEVER 2: UNLOCK LEAPFROG INNOVATION
BY TRANSFORMING R&D WITH AI..........................................................................27
PHARMACEUTICALS..................................................................................................................29
POTENTIAL ENABLERS FOR CONSIDERATION........................................................31
KEY RISKS. ................................................................................................................................32
AUTOMOTIVE................................................................................................................................32
SOFTWARE-ASSISTED VEHICLES (SAVS)................................................................... 33
POTENTIAL ENABLERS FOR CONSIDERATION..................................................35
KEY RISKS............................................................................................................................37
AUTO COMPONENTS DESIGN..........................................................................................38
POTENTIAL ENABLERS FOR CONSIDERATION..................................................38
KEY RISKS............................................................................................................................41
WAY FORWARD FOR THE INDUSTRY................................................................... 41
APPENDIX...................................................................................................................43 10 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
CHAPTER 1: INTRODUCTION
Over the next decade, the adoption of Artificial Intelligence (AI) across sectors is expected
to add $17–26T
1
to the global economy. India’s combination of a large STEM workforce,
expanding R&D ecosystem, and growing digital and technology capabilities positions the
country to participate in this transformation, with the potential to capture 10–15% of global
AI value.
1
Placed against India’s economic outlook, this potential becomes more significant. At its
current growth rate of 5.7%, India’s GDP is projected to reach $6.6T by 2035. However,
under the aspirational 8% growth trajectory outlined in the government’s vision for the nation
known as Viksit Bharat, India’s GDP could increase to $8.3T, representing an incremental
$1.7T compared with the current growth path (Exhibit 1)
AI opportunities for India
Potential AI opportunities for India are presently spread across three levers:
1. Accelerating AI adoption across industries to improve productivity and efficiency,
potentially bridging 30–35% of the gap: Higher output, lower costs of goods and
services, and improved access for underserved markets. These effects are expected to
materialize across both domestic consumption and export markets
2. Transforming R&D, through generative AI, could help India leapfrog into innovation-
driven global opportunities, bridging a minimum 20–30% of the gap: Can generate new
AI-led market opportunities within traditional industries, support commercialization,
reshape legacy value chains, and strengthen long-term competitiveness
3. Innovation in technology services, strengthening India’s reputation as a technology
services leader, contributing another 15-20% to the step up: Could drive the
development of higher-value solutions and new business models, enhancing India’s
competitiveness in the global market
This roadmap focuses on the first two levers, while a separate publication addresses the
third—innovation in technology services.
This roadmap is the first version of this perspective. The insights and recommendations in this
report will be periodically revised, to reflect evolution of the technology as well as the global
economic context. This will keep India’s strategy for accelerated economic development
relevant, resilient and future-ready.
1 McKinsey report titled “The Economic Potential of Generative AI: The Next Productivity Frontier”. June 2023. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 11
Exhibit 1
Exhibit 2
Strategic enablers for AI-led value creation
Realizing the potential of AI in India depends on establishing strategic enablers across
infrastructure, governance, industry, and workforce development. Effective collaboration
between government, the private sector, and academia can support responsible deployment, 12 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
scaling, and skill development, while ensuring broad access and alignment with national
priorities.
• Access to critical AI infrastructure such as cloud platforms, compute, and foundational
datasets could help strengthen India’s sovereign AI capabilities. At the same time,
robust AI governance frameworks, including ethics guidelines and risk controls, could
ensure responsible and secure deployment.
• The private sector can lead the scaling of AI adoption by embedding AI into core
industry processes. This includes driving model validation, secure deployment, and AI-
powered decision-making, while maintaining resilience and accountability at leadership
levels. Reskilling senior executives and upskilling the broader workforce would be key
to enabling this transformation.
• Academia can be vital in anchoring research and supporting large-scale workforce
transformation. The creation of AI testing sandboxes can further enable safety and
scale.
• To ensure inclusive growth, it is essential to provide equitable access to AI resources
and opportunities, particularly for MSMEs and economically underrepresented regions.
These enablers can be potentially mapped into a phased possible path forward, covering
short-, medium-, and long-term priorities aligned with India’s 2035 goals. Progress could be
tracked against established KPIs, with relevant baselines.
India is at a pivotal point in its AI journey. It can capture a meaningful share of AI-driven value
by leveraging its strengths and implementing key enablers. The following chapters explore
the industries, business models, and approaches that could support this transformation.
CHAPTER 2: POTENTIAL OUTCOMES FOR AI-LED VALUE CREATION
AI remains in its formative stage, and market structures are still evolving. To secure a leading
position globally, India could consider investing in sovereign infrastructure, including energy,
to build resilience and unlock higher value-creation potential.
The India AI Mission, with an estimated budget of over ₹10,000 Cr for five years, represents
a foundational step toward strengthening national AI capabilities by focusing on data, talent,
and adoption. Built on seven core pillars, (as of 15th September 2025) the mission plans to
deploy 38,000+ GPUs
2
through a federated compute network, develop India-specific large
language models, and establish an anonymized, consent-based public dataset platform,
placing data at the center as a key enabler for innovation, scalability, and governance in
a diverse, multilingual nation. The initiative also aims to expand AI skilling through the
integration of AI courses at multiple academic levels and creation of AI/Data Labs across Tier
2 and Tier 3 cities, while accelerating adoption through application development in critical
sectors such as agriculture, healthcare, education, and mobility.
2 Press Information Bureau (PIB), Government of India. “Cabinet approves IndiaAI Mission – a significant step towards
boosting India’s AI ecosystem.” March 7, 2024. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 13
While these measures would help lay the groundwork for robust infrastructure, it is also
important that talent pipelines, sectoral uptake, sustained execution, and ecosystem
alignment are in place to realize India’s AI potential. In the following section, the report
identifies potential outcomes that could contribute to the AI-led GDP growth by 2035.
Potential Outcome 1: India becomes the data capital of the world
One of India’s biggest strengths is its data, and it has the potential to leverage this advantage. In
the digital economy, data will function as currency that powers innovation, drives valuations,
and shapes global leadership. India has the potential to lead due to its scale, diversity and
digital infrastructure. By placing quality, trusted, and interoperable data at the core, India
could become the data capital of the world and set new global benchmarks for breadth,
depth, and quality of trusted data ecosystems, potentially through the following options:
• Creating an anonymized data collection framework to easily and safely collect public
data led by entities such as India Data Management Office (IDMO) and National Data
Access Platform
• Building a marketplace of certified non-personalized data with privacy tags and quality
certificate features supported by the National Data Governance Framework
• Developing specialized data platforms for specific sectors as follows:
pFinancial services: Enabling access to cross-industry, alternative data sources with
the IDMO standardizing access and ensuring borrowers retain the right to opt out
pManufacturing: Establishing an open “Manufacturing Data Grid” for OEMs, suppliers,
and startups to trade production and supply-chain data through standard APIs
pPharmaceuticals: Building a unified national omics dataset by sequencing over
10M genomes by 2035 to fuel AI in drug discovery
pAutomotive: Enabling OEMs to share anonymized telemetry data to create a
large-scale sensor dataset for safety and innovation
As of 15th September 2025, AI Kosh, owned and operated by the Government of India under
the India AI Mission, hosts over 2000 curated, non-personal datasets such as census data,
Indian language resources, and satellite imagery.
3
While this is a good foundation, scaling
the breadth and depth of data could position it to move into high-value domains such as
genomics, manufacturing telemetry, and cross-sector financial data, implying that datasets
are certified for quality, tagged for privacy, and interoperable. This can potentially transform
AI Kosh from a foundational national repository into a most trusted and innovation-ready
data platform. India could consider the following options:
• Setting up sector-specific data infrastructure (e.g., for the financial sector) that is
integrated with AI Kosh to host regulatory-grade datasets for model building and
supervision.
• Publishing institution-level AI inventories and publishing a sector-wide AI repository
3 IndiaAI Mission article titled “Now Open: Expression of Interest (EOI) to Contribute Datasets and AI-Artefacts to
AIKosh”. July 7, 2025 14 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
(metadata only) to provide supervisory visibility as AI scales across sectors.
• Integrating AI with Digital Public Infrastructure (DPI) to accelerate inclusive, affordable
financial services at scale (e.g., voice access, Account Aggregator flows, fraud controls).
Potential Outcome 2: India supports the development of an adaptable and efficient
AI-skilling ecosystem
By 2035, India can work towards narrowing the AI skill gap with leading countries by
developing skilled professionals, advancing research, and contributing to AI models. The
focus can be on measuring outcomes of research impact basis international peer-reviewed
publications, the number of PhDs focused on AI research, practical expertise and original AI
contributions based on patents filed. India could consider the following options:
• Harnessing academia better: AI Chairs across the top 20 technology, medical, law, and
business schools to promote and create PhDs/senior qualifications at the intersection
of subject matter expertise and AI
• Incentivizing industry: Finding ways to upskill 3–5% of working professionals through
AI-first modules, supported by tax deductions on employer spending
• Upskilling India: An AI Open University (both physical and online) could enhance Skill
India Digital/eShram in partnership with private ed-tech platforms for the public
• Equipping professionals: Specialized skilling across sectors, illustrated below through
two key examples:
pFinancial services: Launch a national certification program in AI for Credit, Risk
and Fraud, co-designed with top universities
pManufacturing: Initiate a tiered “AI for Advanced Manufacturing” credential in
collaboration with industry leaders
• Exploring AI governance literacy: Look at how to educate Boards and C‑suite of
regulated financial entities covering accountability, risk management, deployment
models (e.g., human-in-the-loop), documentation, and disclosures.
• Ensuring oversight: Consider how a new supervisor entity could help to build capacity
in AI oversight, model risk, audits, and sector risk intelligence.
Potential Outcome 3: Targeted AI adoption unlocks sectoral growth
Focusing on AI enablers across the manufacturing, financial services, pharmaceuticals, and
automotive industries—which represent roughly 25% of India’s projected 2035 GDP—can
help translate AI adoption into measurable outcomes by supporting innovation, improving
productivity, and enhancing export potential. A sectoral analysis highlights these areas as
potential candidates for AI-led innovation and accelerated growth (Exhibit 4).
• Manufacturing: Building world-class smart-factory corridors could enable AI growth
unlocks by: (1) Designating AI-ready industrial parks that co-locate clean-energy
plants with high-performance-computing labs and robotics test zones; (2) Launching
an open manufacturing data grid so OEMs, suppliers and start-ups can exchange real-
time production data via standard APIs; and (3) Rolling out a tiered “AI for advanced AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 15
manufacturing” credential to up-skill engineers from micro-learning to postgraduate
level.
• Financial services: Unlocking responsible AI at scale by: (1) Scaling-up sandbox pilots
to pressure-test explainable credit, fraud-risk and anti-money-laundering models; (2)
Introducing or enhancing existing frameworks to share alternative, consent-based
datasets; and (3) Certifying specialists through a national “AI for Financial Services”
program, creating a trusted talent bench for financial institutions.
• Pharmaceuticals: AI could unlock growth by compressing drug-discovery timelines
and investments by: (1) Expanding biotech parks by 10x and adding best-in-class,
high-performance computing to power AI modelling; (2) Creating a unified national
omics dataset with tiered researcher access; and (3) Training over 100k biopharma
R&D scientists in computational biology and AI by 2035.
• Automotive: Leapfrog to software-led, autonomous mobility by: (1) Setting up six to
eight physical-digital testing parks for India-specific autonomous-vehicle validation;
(2) Deploying 10,000 km of 5G/early-6G “smart corridors” for real-time vehicle-to-
infrastructure data flow; and (3) Mandating anonymized telemetry from 20–25% of
new vehicles each year to seed national safety and innovation datasets.
This analysis could be applied to construction, wholesale and retail trade, and professional
services, contributing another 25% of the projected 2035 GDP and ensuring AI-driven value
scales across the economy (Exhibit 4).
Potential Outcome 4: India’s jobs are future-proofed, and industry transformed at scale
India can address fragmented skilling through a unified system that enables continuous
worker upskilling, accelerates firm-level digital adoption, and strengthens safety nets. This
would mean mapping job shifts annually, embedding lifelong learning into career pathways,
scaling MSME digital upskilling, and protecting gig and platform workers – projected to reach
about 23.5M by 2029–30
4
. International benchmarks highlight both the urgency and the
opportunity: the World Economic Forum estimates 23–25% of roles will change within five
years (69M created; 83M disrupted)
5
, the OECD finds 27% of jobs are in occupations at high
risk of automation
6
, the U.S. may see 12M occupational transitions by 2030 as generative
AI scales
7
, and earlier global studies suggest China could face up to ~100M such transitions
under fast automation scenarios
8
. India could consider the following options:
4 NITI Aayog policy brief titled “India’s Booming Gig and Platform Economy: Perspectives and Recommendations on
the Future of Work”. June 2022
5 Future of Jobs Report 2023 titled “Future of Jobs Report 2023: Up to a Quarter of Jobs Expected to Change in
Next Five Years”. April 30, 2023.
6 OECD report titled “OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market”, finding that
approximately 27% of jobs across OECD countries are in occupations at high risk of automation (including AI). July
2023
7 McKinsey Global Institute report titled “Generative AI and the future of work in America”. July 26, 2023.
8 CSIS ChinaPower Project analysis titled “Is China Ready for Intelligent Automation?” August 25, 2020 16 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• Continuous reskilling: Developing job transformation maps for 25-30 priority sectors
can help to identify task shifts, emerging roles, and reskilling pathways, drawing on
models such as “Workforce Singapore”
9
. Large enterprises could prepare and submit
firm-level skilling plans.
• Accessible learning: One approach could be introducing digital, portable individual
learning accounts. Singapore’s “SkillsFuture”
10
and the UK’s “Lifelong Learning
Entitlement”
11
offer interesting models to consider. These accounts could be designed
to link credit reimbursements to verified course completion and employment outcomes.
• Industry-wide AI adoption: By preparing industry AI plans, with curated toolkits, vetted
vendor lists, and sector-specific micro-credentials. Consider Singapore’s “Industry
Transformation Map” framework as a reference
12
.
• Supporting gig and platform workers: By implementing the Code on Social Security
(2020)
13
, ensuring universal registration and benefits such as health insurance, accident
cover, and retirement savings that can be carried between jobs. Consider using the
e-Shram portal to manage benefits and pilot wage-loss insurance schemes for workers
facing job loss.
• Supporting at-risk worker groups, a live, integrated skills and jobs database by
linking e-Shram, Skill India Digital, and public job listings into a constantly updated
system that tracks in-demand skills, identifies at-risk worker groups, and connects
them directly to funded training and verified job opportunities.
• Financial consumer and worker AI literacy. Ensuring disclosures and grievance
pathways into adoption programs so that trust leads to usage.
Guided by these potential outcomes, the report proceeds to examine two primary AI
opportunity levers for India: Lever 1 focuses on accelerating AI adoption across industries
to enhance productivity and efficiency, while Lever 2 explores the transformation of R&D
through generative AI, enabling India to capture innovation-driven opportunities and bridge
a significant portion of the growth gap.9 Workforce Singapore website
10 SkillsFuture Singapore (“SSG”) government website
11 House of Commons Library research briefing titled “The Lifelong Learning Entitlement.” Published 12 March 2024.
12 Singapore Ministry of Trade and Industry webpage titled “Overview” (Industry Transformation Maps under the S$4.5
billion Industry Transformation Programme). Last updated September 8, 2025
13 Gazette of India (Extraordinary), titled “The Code on Social Security, 2020 (No. 36 of 2020)”. 28 September 2020. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 17
CHAPTER 3: LEVER 1 - ACCELERATING AI ADOPTION ACROSS INDUSTRIES
TO IMPROVE PRODUCTIVITY AND EFFICIENCY
To assess AI’s potential for India, a detailed analysis was conducted on its ability to enhance
productivity across industries. The study covered over 850 occupations across 16 sectors and
examined more than 2,100 distinct work activities. The analysis indicates that AI adoption
could contribute an additional $500–600B to India’s GDP by 2035, beyond the projected
growth trajectory, driven by productivity improvements, operational efficiencies, and the
reallocation of human effort to higher-value tasks.
The following sections detail the methodology and findings behind these projections.
Approach
Specific adoption scenario models were considered (i.e., the pace at which industry adopts
the technology at scale, resulting in impact on productivity)—early, midpoint, and late—to
estimate when AI could effectively take on these activities based on currently demonstrated
technologies and their expected development in the future and country-specific factors such
as wage levels and occupational mix. The model incorporates software capabilities such as
machine learning, data analytics and hardware-driven automation such as robotics.
• Baseline employment and GDP data: Used 2022 as the baseline year for both
employment and real GDP, sourced from IHS. Calculated productivity as GDP per
worker to set the reference point for future projections
• AI adoption rates for 2035: Estimated sector-level AI adoption rates using McKinsey
Global Institute’s (MGI) model, covering about 850 occupations and 2,100 activities
with sectoral nuances. Applied these AI adoption rates to baseline employment to
determine workforce segments likely to be automated
• Growth rates across different scenarios: Augmented workforce calculated by
redeploying the automated workforce at current productivity levels. New GDP projected
by applying 2022 productivity to the augmented workforce. This yielded GDP CAGR
over 2022–2035, forming the basis of Lever 1 sectoral projections
• Relevant scenarios chosen (illustrated below in Exhibit 3): For each sector, AI
adoption across early, mid, and late horizons, under two cases were modeled- AI
adoption scenarios across late, mid and early for each sector. Acceleration is assumed
at two levels, leading to two scenarios: the accelerated AI adoption scenario assumes
faster tech adoption with sectors shifting to earlier phases by 2035, while the moderate
scenario assumes slower adoption and later starting points
• GDP 2024 and business-as-usual CAGR 2024-2035: Estimated sector-wise data from
the IHS database, extrapolated to align with government projections for a total of
$3.6T with sectoral nuances
• Final incremental AI productivity impact value: Derived using an additional productivity
boost to the expected business-as-usual CAGR and then applying it to the GDP 2024
numbers to receive GDP with AI adoption 18 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Assumptions
Below are the key assumptions that have been considered in the model (Exhibit 3):
• The model evaluates current capabilities required to automate tasks in each sector
based on their complexity and AI readiness
• Technologies are assessed on whether they can match or exceed the performance of
top-quartile human workers for specific tasks
• Adoption considers the estimated time to build viable AI solutions and whether they
are economically justified from a cost-benefit perspective
• The degree of AI adoption in a sector is closely linked to its occupational structure (i.e.,
task types and automatable roles)
• Displaced workforce due to AI is assumed to be redeployed within the same sector,
maintaining existing sectoral labor productivity levels
• While India has historically seen slow AI adoption, sector-specific improvements in
digital infrastructure and technology maturity can enable faster adoption going forward
• Adoption trajectories differ by sector based on readiness of solutions, implementation
lag, and cost-benefit feasibility
• Productivity gains are expected to be redeployed at similar levels or with a reduction
of up to 20%
Exhibit 3
Results
The results indicate that accelerated adoption of AI across industries can contribute
$500B-$600B over and above India’s current GDP growth by 2035, driven by increased
productivity and efficiency in the workforce (Exhibit 4). The analysis shows that financial AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 19
services and manufacturing can be most impacted and might have up to 20-25% of their
sectoral GDP attributed to AI by 2035. Both sectors are detailed in the sections ahead.
Exhibit 4
Banking
AI-led productivity and efficiency improvement could unlock $50B-$55B in financial services,
over and above the current estimated growth for the sector by 2035 (Exhibit 4). This opportunity
will likely be realized as AI-mature Indian banks evolve into “bionic” organizations, combining
machines’ intelligence with humans’ judgment. AI accelerates business and functional
transformation across the banking value-chain, embedding intelligence into every product,
process and customer interaction. Reducing costs can also enhance financial inclusion.
Financial services companies’ front, middle and back offices are expected to be transformed
by machine learning and agentic AI. While the map represents important opportunities across
domains, areas with potentially the highest ROI have been highlighted (Exhibit 5):
• In the back office, AI could power automated compliance, fraud detection, and risk
management through advanced anomaly detection techniques and privacy-preserving
analytics such as secure multi-party computation and federated learning
• In the middle office, AI-enabled systems can reshape credit decisioning, collections,
and portfolio management. By leveraging alternative data sources, banks can make
more accurate, dynamic, and inclusive lending decisions
• In the front office, virtual relationship managers can deliver hyper-personalized
customer experiences. Using real-time behavioral predictions, these AI agents can offer
tailored financial advice, timely product recommendations, and proactive outreach,
helping deepen customer engagement and improve satisfaction across segments 20 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 5
...and a combination of AI & machine learning can transform financial services across
the front, middle, and back office
Front office Middle office Backoffice
Digital-led customer acquisition
Adaptive look-alike models scan daily click-
stream, score prospects, and auto-shift ad
budgets toward the highest-conversion
channels, boosting CAC efficiency.
Frontline sales enablement
Real-time call co-pilot transcribes the conversation,
matches needs to product bundles, inserts
mandatory compliance language, and logs next
steps straight into the CRM.
Relationship management and advisory
Generative assistant assembles concise
portfolio snapshots, flags risk or life-event
triggers, and drafts personalized action plans
for the relationship manager (RM) to approve
and send
Engagement, cross-selling, and customer
retention
Life-stage engine analyses transaction patterns and
sentiment to predict churn or upsell windows, then
launches hyper-targeted offers via push, email, and
RM dashboard.
Customer underwriting
Explainable ML combines bureau, cash-flow,
utility, and GST feeds to deliver real-time
affordability scores with clear reason codes
for the credit officer and the regulator.
Collections
The early delinquency model prioritizes overdue
accounts, picking the best channel, timing, and
repayment offer to maximize recovery at the lowest
collection cost.
Self-service through digital channels
Multilingual chatbot authenticates with
biometrics, handles KYC updates, disputes,
and card blocks end-to-end, and escalates
only edge cases to a human agent.
Assisted service
(contact center, branch, digital)
Voice analytics gauges sentiment and intent mid-
call, suggests relevant knowledge-base snippets,
while back-office bots auto-populate and route
service tickets.
Developer productivity
AI pair-programmer auto-writes routine
code, adds unit tests, tunes SQL queries,
checks for security issues, and flags build
problems before the code is merged. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 21
Potential enablers for consideration
Realizing the full potential of AI in India’s banking sector will depend on a set of enablers that
support innovation, adoption, and responsible scaling. These include:
Infrastructure
• Building capacity through Innovation Sandboxes to enable pilots focused on critical
themes such as explainable credit models, fraud and AML graph analytics, and self-
auditing regulatory technologies
• Utilizing a regulatory sandbox to test AI-related regulatory changes, e.g., video KYC
for NRI.
• Harness open, standardized dashboards from pilots to track business impact, fairness
outcomes, and emerging risks, enabling transparent supervision and learnings across
the ecosystem
• RBI has already announced the establishment of a public cloud infrastructure for the
financial sector
14
. Such efforts can be accelerated:
pAn empaneled set of vendors offering high-performance compute (e.g., GPUs),
privacy-preserving tools, and LLM-based APIs
pStrong data protection frameworks, including data replication, disaster recovery,
and compliance checks
pPre-approved toolsets for common banking use cases, made easily accessible
through a curated marketplace
• Define standards for and enable access to utility agents, trained on regulatory-grade
datasets and made available to banks and other financial institutions. RBI can define
standards to ensure these agents are explainable, compliant, and regularly updated to
reflect regulatory changes and market evolution
• Harness a cross-regulatory AI Innovation Sandbox to enable financial institutions to
test models in a secure environment alongside the regulatory sandbox.
• A shared “landing zone” for GPUs and computing resources on a pay-per-use basis for
smaller regulated entities, delivered via RBI or sector infrastructure providers.
• A dedicated funding corpus could support shared data and compute as public goods,
including grants for fintech accelerators and research labs.
• A sector-specific AI model for finance and make them available for safe adoption
across the sector.
14 Reserve Bank of India. Monetary Policy Statement. December 8, 2023. 22 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Data
• A central regulatory body for data governance in the financial sector. It could include,
but not be limited to, defining standards for classification of data, sharing across
entities, responsible use, security, and monetization.
• Defining frameworks (either enhancing existing or developing new ones) to include
alternative (especially unstructured) data sources, under a unified consent and
governance structure. An adequately authorized entity, such as the India Data
Management Office (IDMO), can standardize look at access issue to ensure borrower
opt-outs, and enable explainable scoring aligned with regulatory norms
• Harness existing data architecture frameworks to facilitate secure, anonymized data
exchanges between banks and other financial institutions for the purpose of model
training, ensuring privacy and transparency
• Broader adoption of alternative data by banks through enabling policies and common
infrastructure. An adequately authorized entity, such as the RBI, could:
pIssue clear guidelines on the use of such data for credit and other financial decisions
pEstablish a shared, consent-driven data repository, accessible to regulated entities
pPromote standardization and interoperability for secure data access and validation
• Consider whether a financial institution could maintain a regular inventory of its AI
systems, with anonymized metadata fed into a sector-wide repository for supervisory
visibility.
• AI can be more deeply integrated with existing digital public infrastructures (e.g., UPI,
AA, OCEN) to support enhanced public services such as multilingual access and fraud
detection.
Talent
• Training programs focused on the needs for the financial sector. As an example, a
national certification program in AI for Credit, Risk and Fraud could be considered
by the government, industry organizations, and universities. This would build trusted
and domain-aligned talent with expertise in responsible AI, explainability, and financial
regulation
• Consider whether tax incentives on course fees for financial institutions could increase
investment in skilling employees in AI, analytics, and cybersecurity. This would reduce
the post-tax cost for employers and spur large-scale adoption of upskilling programs
• A scalable and flexible talent pool through a national AI fellowship or exchange
platform:
pMaintain a vetted database of certified AI professionals available for temporary
deployments across banks, and other financial institutions, including regulators
pAllow small and mid-sized institutions to access top talent on-demand without full-
time hiring overheads AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 23
pFacilitate secondments or rotational programs across banks, RBI innovation arms,
and academia to deepen cross-sectoral expertise and rapid capacity building
Protection, Assurance and Governance
• Consumer protection and transparency by mandating fairness in AI outcomes, clear
disclosures on AI use, and accessible grievance redressal mechanisms, supported by
public reporting and toolkits that help smaller firms meet compliance.
• Resilience and security through continuous cybersecurity monitoring, dynamic threat
response, regular red-teaming of AI systems, and business continuity plans with
fallback mechanisms and drills.
• Governance and oversight by requiring AI-specific risk checks in product approval
processes, continuous monitoring of deployed models, institution-level AI inventories
feeding into a sector-wide repository, and risk-based audits (with independent third-
party audits for high-risk systems).
• Accountability through incident reporting and risk intelligence by establishing
tolerant, good-faith reporting mechanisms and aggregating disclosures to generate
sector-level insights on emerging risks.
Key risks
While AI in banking offers significant promise, India must also navigate a set of risks that
could impede adoption.
Data Infrastructure Talent Regulatory and IT Policy
Legacy IT core infrastructure
Most banks still operate on bulky legacy core
platforms. Integrating AI tools often requires
extensive rework or replacement, adding cost
and multi-year delays to realize productivity
benefits. These delays will persist without
sovereign AI/cloud infra and curated toolsets.
Data coverage and quality gap
Despite the push for Account Aggregator (AA) and
DEPA frameworks, many individuals remain outside
the digital data net, especially in low-digitization
areas. A lack of standardized and consent-driven
repositories further limits the availability and
reliability of alternative data for AI models.
Sandbox throughput constraints
Unlock the Innovation Sandbox’s capacity to
evaluate important interventions along the
financial services value chain. Without open
dashboards and dissemination of learnings,
systemic benefits will remain limited.
Privacy and consent fatigue
Consumer privacy concerns around alternative data
are intensifying. Without robust data protection,
explainability, and opt-out frameworks, public trust
may erode, constraining data access.
Lack of scalable AI utility infrastructure
Absence of centralized AI tools, agents,
and vetted vendors for common banking
use cases forces every bank to reinvent the
wheel. This fragmentation hinders innovation
velocity and increases cost.
Unequal access to AI talent and tools
Smaller banks and fintechs may lack the resources
to tap into centralized utility agents, high
performance compute environments or rotational
AI pools, reinforcing a digital divide within the
financial sector.
Upskilling & change management
Without clear roles, staff training, and
guardrails, using AI agents can lead to errors,
compliance issues, and serious risks to
customer trust and system stability. 24 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Manufacturing
In manufacturing, $85–100B could be driven by AI-led productivity and efficiency improvement
over and above India’s current growth by 2035. The National Manufacturing Mission outlines
five key pillars
15
: Ease of Doing Business, Future-ready Workforce, Vibrant MSME Sector,
Availability of Technology, and Quality Products, of which AI will have a high impact on three:
Availability of Technology, Future-ready Workforce, and Vibrant MSME Sector.
AI can unlock productivity and efficiency across multiple dimensions by lowering the cost
of production, improving output yields through enhanced process efficiency, and increasing
throughput via predictive maintenance on the shop floor. It can also enable the production
of higher-quality goods at similar prices by powering intelligent product design, real-time
quality control, and mass customization. To fully realize these benefits and build a future-
ready, competitive industrial base, upskilling India’s manufacturing workforce in AI tools will
be essential (Exhibit 6).
For India to fully capture the gains from AI-native manufacturing, it is important to strengthen
both forward and backward linkages. On the backward side, this means building resilient
supply chains, integrating AI-ready MSMEs, and ensuring reliable access to inputs. On the
forward side, India can actively expand domestic markets and position itself in global value
chains through coordinated industrial and trade policies. Productivity gains alone will not
deliver impact unless industrial policy, trade strategy, and demand generation evolve together
to convert efficiency into competitiveness and growth.
Exhibit 6
15 Press Information Bureau, Government of India. “‘National Manufacturing Mission’ to cover small, medium and large
industries for furthering ‘Make in India’ announced in Union Budget 2025-26.” February 1, 2025 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 25
Potential enablers for consideration
Realizing AI’s full potential in India’s manufacturing sector will depend on enablers that
enhance productivity and foster innovation-driven growth at scale.
Infrastructure
• “AI-Ready Industrial Parks” that place clean-energy factories next to high-performance
computing (HPC) labs and skills centers, similar to the technology-and-fabrication
clusters promoted under the United States CHIPS and Science Act
• Government decision makers to look at -funding shared facilities for 3-D printing,
advanced materials testing and precision metrology can give MSMEs affordable access
to expensive equipment
• Robotics and humanoid test zones with safety rules that allow rapid prototyping,
already being done in several Chinese electric-vehicle (EV) hubs, could speed up local
innovation in collaborative robots and vision-guided welding stations
• A national BharatNet fibre backbone with Time Sensitive Networking (TSN) - capable
switches will let control signals travel reliably between separate buildings or even
distant factories that share the same production line
• An online SaaS marketplace where Indian and global developers can publish plug-and-
play AI tools, for example, for anomaly detection or energy optimization, which will let
factories subscribe to the exact algorithms they need instead of investing heavy capital
upfront
Data
• An open “Manufacturing Data Grid,” a shared platform where OEMs, suppliers, other
stakeholders, and startups can trade production and supply-chain data through
standard APIs, taking inspiration from Germany’s Manufacturing-X data-space model
• Industrial parks to house both cloud servers and edge computers equipped with GPUs
so that high-speed quality checks, digital twins and predictive-maintenance systems
can run close to the machines that generate the data
• Industrial clusters with 5G and 6G mobile networks combined with Time-Sensitive
Networking (TSN) switches so robots and Automated Guided Vehicles (AGVs) can
send and receive data without delay
• Data sharing framework for manufacturing so that companies can share sensitive
design drawings or process settings only with trusted partners and always with clear,
revokable permissions
• National libraries (or marketplaces) of reusable digital-twin models, for example, for
semiconductor fabrication plants, battery-cell lines, precision auto-parts machining
and aerospace composites to shorten design cycles without starting every simulation
from scratch
Talent
• The Ministry of Skill Development, the All-India Council for Technical Education
(AICTE) and leading manufacturers could explore how to launch a tiered “AI for
Advanced Manufacturing” program that starts with micro-badges and scales up to full 26 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
postgraduate programs, following Japan’s Ministry of Economy, Trade and Industry
(METI) model for rapid AI upskilling
16
• Industry-academia research chairs in specific domains, e.g., chip design, power
electronics, robotics and battery chemistry to allow professors and graduate students
to work on real-world factory challenges while keeping intellectual property inside
India
• Government decision makers could assess the benefits of a targeted reverse-diaspora
program that offers fast-track visas, research grants and senior leadership roles to
encourage experienced semiconductor, battery and automation experts living abroad
to return and teach or build in India
• A national registry of certified freelance AI-maintenance engineers, digital-twin
modelers and automation troubleshooters that will help factories scale their specialist
workforce up or down
Key risks
Risks that could hinder India’s push toward AI-native manufacturing include
Data Infrastructure Talent Regulatory and IT Policy
Fragmented data ecosystem
Without a standardized Manufacturing Data
Grid and common APIs, production and
supply-chain data will remain siloed across
OEMs, MSMEs, and startups. This limits AI’s
ability to deliver end-to-end visibility and
optimization.
Inadequate edge + network infrastructure
AI use cases like digital twins or predictive
maintenance need low-latency computing close
to machines. Many parks lack edge GPUs, 5G/6G,
or TSN-capable networks slowing adoption.
Shortage of cross-skilled talentShortage of cross-skilled talent
India’s AI workforce has room to deepen
expertise in manufacturing (e.g., robotics,
chip design, battery tech). Upskilling via tiered
credentials and reverse diaspora programs
is critical but currently limited in scale and
alignment with real factory needs.
Low awareness and access to shared AI
infrastructure
MSMEs often can’t access HPC labs, 3D printing
centers, or plug-and-play SaaS AI tools due to
cost, geography, or lack of knowledge, widening
the gap between large OEMs and small suppliers.
Slow adoption of DEPA-like consent
protocols
Sharing sensitive factory data (e.g., designs,
process settings) is essential for collaborative
AI. Without robust consent-based, revokable
frameworks, companies may avoid data
sharing due to IP fears.
Slow industry readiness tracking
Without government support and structured
assessments like SIRI, many MSMEs and smaller
organizations may struggle to identify critical
skill or technology gaps—slowing their ability to
scale and transition toward Industry 5.0.
AI adoption across industries represents a critical lever for India to enhance productivity
and competitiveness, with banking and manufacturing standing out as early opportunities.
Building on these foundations, the next chapter explores Lever 2: Unlocking leapfrog
innovation by transforming R&D with AI—a pathway for India to accelerate discovery, shorten
innovation cycles, and establish a stronger foothold in global, innovation-driven industries.
16 Ministry of Economy, Trade and Industry (METI), Approaches to human resources and skills required for DX
promotion in the age of generative AI, August 7, 2023. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 27
CHAPTER 4 – LEVER 2: UNLOCK LEAPFROG INNOVATION BY
TRANSFORMING R&D WITH AI
Transforming R&D, especially using generative AI, can enable India to leapfrog into innovation-
based opportunities on a global scale by transcending traditional growth pathways and
creating new products, services, and business models. Historically, such innovation paths have
been challenging for India due to the heavy capital required for conventional R&D, e.g., $1–2B
17
for novel drug development and $1–1.5B to engineer an all-new electric vehicle platform. AI
lowers these entry barriers. It can accelerate commercialization, disrupt legacy value chains,
and create a lasting competitive edge. Analysis suggests that such breakthrough innovations
could potentially contribute at least an incremental $280-475B to India’s GDP by 2035.
Approach
11 sectors have been selected from McKinsey’s 18 global arenas
18
of growth and 7 additional
India-specific arenas based on their local relevance and transformative potential.
• Baseline (2023) revenues across all 18 arenas were sourced from current industry reports
and total $640–750B
19
. Projections for 2030 were developed using a combination of
industry estimates and expert consultations (e.g., ONDC for next-gen e-commerce),
reaching $1.7–2.0T
• The 2035 revenue projections were developed using 2023–2030 growth trends,
resulting in an estimated $3.6–4.5T range. To assess the incremental GDP contribution,
the revenue delta from 2023 to 2035 was calculated and sector-specific revenue-to-
GDP conversion factors were applied, resulting in a projected GDP impact of $1.4–1.9T
across all 18 arenas by 2035
• Out of 18 arenas, 10 were identified where AI is the primary value driver for leapfrog
growth. These 10 arenas are expected to contribute $380–660B in incremental GDP
impact by 2035
• This yields an estimated $280-475B in incremental GDP attributable specifically to AI-
led leapfrog growth across the 10 arenas by 2035
The exhibit below shows the identified 18 arenas of growth that could serve as key drivers of
India’s growth over the next decade, with a projected GDP impact of $1.4–1.9T by 2035.
17 World Economic Forum article titled “How blockchain can cut the cost of new medicine”. December 2018.
18 McKinsey Global Institute report titled “The Next Big Arenas of Competition”. October 23 2024 – 18 global arenas of
which 11 are directly relevant for India and 7 additional were added
19 McKinsey article titled “India’s Future Arenas: Engines of Growth and Dynamism”. June 19 2025. 28 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 7
Of the 18 arenas of growth, ten are primarily driven by AI-led innovations such as AI software
and services, a global hub for auto-components, cloud services, software-assisted vehicles,
semiconductors and PCBs, medical devices, SpaceTech, aerospace and defense, cybersecurity,
and biopharma (see exhibit below).
Exhibit 8
It is important to recognize that AI-led growth will be shaped by unexpected breakthroughs
that remain beyond our foresight today. AI is advancing at an extraordinary pace with AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 29
AI supercomputing capacity doubles every 8–10 months.
20
, and AI algorithmic efficiency
doubles every 15–18 months
21
. Given this velocity, it is difficult to anticipate the full spectrum
of innovations that may arise. While 18 opportunity industries for India have been identified,
the rapid evolution of AI could well unveil a 19th or 20th frontier that will likely emerge over
time.
The next section illustrates the AI-led opportunity in the pharmaceuticals and automotive
sectors, which are aligned with India’s factor endowments.
Pharmaceuticals
Currently, 80% of the Indian pharmaceutical market is driven by generics
22
. This is because
the high costs of developing a novel drug (up to $1–2B per molecule)
23
, long timelines (over
10 years)
24
, and significant financial risks have historically limited investment in innovative
R&D capabilities. Emerging technologies such as AI can help lower development costs and
timelines across the drug discovery and development value chain, enabling India to transition
from generics to the innovator space over the next decade. India’s expertise in generics,
domain talent (e.g., pharmacology) and its endowment in the form of a rich genetic pool can
position it well to capture this opportunity.
Traditional drug development is divided into five distinct stages and typically takes >10 years
to complete end-to-end, with a potential capital spend of $1–2 billion (Exhibit 9).
Exhibit 9
20 arXiv preprint titled “Trends in AI Supercomputers”. April 22 2025
21 Forethought Research article titled “Will AI R&D Automation Cause a Software Intelligence Explosion?” March 25
2025
22 IQVIA market data
23 World Economic Forum article titled “How blockchain can cut the cost of new medicine”. December 2018
24 N-SIDE blog post titled “What’s the average time to bring a drug to market in 2022?”. November 5 2022 30 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
AI is re-shaping the traditional drug discovery value chain. It could reduce R&D costs by
20–30%
25
through drug repurposing, AI-driven research and documentation, and replacing
traditional placebo groups in clinical trials with AI-generated virtual placebos. This can
simulate control groups without needing real participants; shorten drug discovery timelines
by 60–80%
26
via AI-powered molecule design and Insilico modelling to speed up lead
identification by four times; improve clinical trial success rates by 5–15% by leveraging India’s
diverse gene pool to identify optimal patient subgroups.
Additionally, frontier technologies are creating new pathways for drug development. Platform-
based approach is enabling the creation of reusable, end-to-end technology engines that
can discover, design, and optimize multiple drugs across therapeutic areas by integrating
AI models, large-scale genomic data, and automated lab systems. One key advantage of
this approach is the ability to repair or retune existing molecules in days or weeks when a
pathogen or tumor mutates, avoiding the need to restart the discovery process from scratch
For example, a Boston based startup advanced a novel lung fibrosis drug from concept
to Phase 1 trials in under 30 months, while another Canada-based startup developed a
COVID-19 antibody in just 90 days using its AI-driven antibody discovery platform.
India could consider licensing and launching 90–110 innovative drugs by 2034 across four
phases: post-molecule discovery, post-phase 1, post-phase 2, and E2E commercialization.
This would culminate in a value capture of $5–8B and establish India as an innovation-led
hub (Exhibit 10).
Exhibit 10
25 McKinsey report titled “Generative AI in the pharmaceutical industry: Moving from hype to reality”. January 9 2024
26 McKinsey report titled “Generative AI in the pharmaceutical industry: Moving from hype to reality”. January 9 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 31
Potential enablers for consideration
Realizing value in pharmaceuticals could depend on enablers to improve clinical research,
optimize manufacturing and ensure regulatory readiness.
Infrastructure
• Expanded the number of biotech parks by 10x to support the predicted expansion in
research and development led by the growth of startups
• High-performance computing infrastructure in line with EU and China to meet the
intensive computational needs of AI-driven drug discovery
Data
• Diverse genomic and clinical datasets to build a unified, high-quality national omics
dataset, led by institutions such as the Indian Biological Data Centre (IBDC) and
Biotechnology Industry Research Assistance Council (BIRAC)
• A tiered data access model like the U.S. National Institutes of Health’s “All of Us”
program, providing public, registered, and controlled access levels to researchers
27
Talent
• Train and retain biopharma R&D scientists by 2035 to support the annual development
of 20–25 new drugs (Exhibit 10)
• Undergraduate and postgraduate programs in computational biology and bioinformatics
in collaboration with premier technical institutes
• Expanded central research grant pools to have a dedicated track for AI in drug discovery
that focuses on attracting global talent
• Enhanced re-entry fellowships with increased stipends and research grants to match
global talent programs
Policy and Regulations
• Government decision makers could review India’s pharmaceutical regulations to ensure
they align with global standards to facilitate faster clinical trials and international
recognition
• Government decision makers could explore the potential benefits of a data exclusivity
law that could protect clinical trial data while incentivizing innovation
• Government decision makers could explore whether there is scope to streamline the
clinical trial approval process, a potential 30-day approval route for institution-initiated
trials would match global best-in-class timeframes
• Government decision makers could look at how to implement global best practices
for vaccine approvals, such as rolling data reviews and digital submissions that aim to
shorten overall timelines for time to market
27 All of Us Research Hub article titled “Data Access Tiers” 32 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Market Access
• A drug-access acceleration fund through BIRAC to co-invest in the commercialization
of Indian therapies
• A “Made-in-India Innovation Track for Pharma” to mandate the listing of breakthrough
drugs within 180 days of approval
• Empower pre-revenue biotech firms to go public by mirroring NASDAQ’s non-revenue-
based listing parameters
• Consider the potential benefits of offering capital incentives, such as capital grants for
AI-R&D centers and reduced tax rates on profits from India-patented drugs
• Global pharmaceutical firms with priority access to national high-performance
computing infrastructure and a 180-day fast-track approval path for AI-designed drugs
Key risks
India’s goal of AI-based drug development faces potential risks
Data Infrastructure Talent Regulatory and IT Policy Market access
Value erosion
U. S . market reforms could cut global drug out-
licensing value
28
, sharply curbing the earnings
potential of Indian-origin therapies.
Slow and uncertain domestic market
Unless the pharma regulatory body shortens
drug-approval timelines from 20–25 months
to 15 months, matching global benchmarks,
each year of delay can erode 10-15% of an
asset value.
29
.
Unclear guidelines on AI-discovered drugs
Global patent norms requiring “significant
human contribution” may limit protection for
AI-generated molecules, risking rejection and
long-term barriers to global commercialization
in key markets.
Genomic data gap
Reaching the omics dataset target by 2035
faces major risks, high capex, potential legal
backlash over consent, and low participation
due to rising privacy concerns.
High-performance compute bottleneck.
AI in drug discovery requires dedicated HPC
capacity, but risks like lack of funding, global
GPU shortages, 70+ week lead times
30
, and
export controls could delay deployment by
18–24 months.
R&D talent gap
India’s low researcher density and
uncompetitive stipends may hinder progress
toward the 2035 target of 1.5k researchers per
million, unless talent training and retention
improve.
Automotive
AI can emerge as a game changer for the automotive sector in India, enabling it to cut
costs, improve safety and accelerate innovation. In the following sections, the report
explores two pathways for automotive: Software-Assisted Vehicles (SAVs) and AI-enabled
component design. Harnessing frontier technologies, including RFID-based smart corridors,
5G-connected routes, and AI-driven design and validation, could put 18-20M software-
ready vehicles on Indian roads by 2035 and unlock $20-25B in export gains and import
substitution. (Exhibit 12 and Exhibit 14)
28 OHE bulletin titled “US drug pricing policies will have global impacts on innovation and access”. June 30 2025
29 McKinsey article titled “The road to positive R&D returns”
30 Industry Technology Report titled “Prepare for the Coming AI Chip Shortage”. September 25, 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 33
Software-assisted vehicles (SAVs)
Software-Assisted Vehicles (SAVs) represent the next generation of automobiles, where
core functionalities are increasingly driven by software rather than hardware-intensive
systems. SAVs operate across five defined levels of autonomy, as per the Society of
Automotive Engineers (SAE) International (Exhibit 11). The automotive industry in India
is presently concentrating its efforts on progressing from Level 2, which features partial
vehicle autonomy, towards achieving Level 4, characterized by highly autonomous driving
capabilities. These vehicles rely on flexible electronic architectures, connected systems, and
over-the-air (OTA) updates to minimize human intervention.
India is expected to reach Level 3 by 2035, with its AI-led automotive inflection point
between Levels 3 and 4. As India emerges as a major SAV consumer market and global
production hub, this shift offers a key opportunity for domestic value creation and global
competitiveness.
Exhibit 11
By 2035, 40-50% of the total 40-45M vehicle market, i.e. 18-20M units, would be enabled
via software (Exhibit 12). These will be split across passenger vehicles at 4-5M, commercial
vehicles at 1-2 M and two-wheelers at 13-15M. India could unlock $6-8B cumulative domestic
value through AI-enabled SAV domestic subscriptions by 2035, with an estimated $1.5-2B
annual exit value that year. 34 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 12
A three-level autonomy-driven pathway for India’s SAV evolution is considered. India’s SAV
evolution will move from Level 2 partial automation (driver-assisted) to Level 3 conditional
automation (hands-off in specific scenarios) and Level 4 high automation (self-driving in
geofenced areas).
Unlocking opportunities beyond conventional technologies
Conventional autonomous driving relies on costly on-board sensors like LiDAR, cameras,
and radars combined with real-time AI processing to navigate roads without human input.
However, such systems may face challenges in India due to weather or traffic conditions
and poor road markings. India can explore alternate, infrastructure-assisted approaches to
enable affordable and reliable autonomy, including:
• RFID-based corridors that help vehicles localize accurately in all weather, reducing
reliance on GPS or cameras
• Magnet marker guides lanes, which act like a virtual rail, enabling lane-keeping, even
on waterlogged roads
• 5 G-enabled corridors that share real-time data over 5G, helping vehicles detect hidden
traffic in urban settings
• Satellite navigation system in combination with real-time kinematic from ground towers
provides centimeter-level accuracy ideal for remote areas (Exhibit 13). AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 35
Exhibit 13
To unlock $6-8B in SAV value over the next decade, AI can help accelerate engineering
capability building and faster time to market for software-enabled automotive products at
reduced costs. AI copilots or custom LLMs trained on AUTOSAR/ vehicle OS documents
can cut learning time by 30–50%. AI-powered model-based design can reduce design cycle
times by 20–30%, while AI-led bug detection and regression testing can lower validation
costs by up to 40%. Additionally, AI vision and reinforcement learning can optimize compo-
nent assembly and reduce Electronic Control Unit (ECU) testing costs.
Potential enablers for consideration
Realizing the full potential of shared autonomous vehicles in India will depend on enablers
that build robust digital infrastructure, ensure safety and regulatory readiness.
Infrastructure
• 6–8 physical-digital testing parks
31
by 2035 to validate autonomous driving features in
Indian conditions
• 10,000 km
32
of 5G/early-6G corridors with roadside units to enable real-time data
exchange for software-assisted vehicles
• Designated 10-20 low-risk deployment zones by 2030 in areas such as airports and
campuses for the initial commercial rollout of Level 3 and Level 4 autonomous vehicles
31 Benchmarking with testing parks in South Korea and U.S. per million cars; American Center for Mobility report titled
“About the American Center for Mobility”; K-City report titled “K-City: South Korea’s 5G-Connected Autonomous
Vehicle Testbed”
32 Covering Golden Quadrilateral, connecting Delhi, Mumbai, Chennai, and Kolkata (~5,900 km); high-impact national
corridors (~2,000 km), including Ahmedabad–Nagpur, Mumbai–Bangalore, and Delhi–Kolkata; and top 20 urban
ring roads (~2,000 km) 36 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• Three regional Zero Prototype Labs in key automotive component hubs e.g., Pune,
Aurangabad that provide digital testing and simulation facilities, particularly for MSMEs
and startups
Data
• Incentivized manufacturers sharing anonymized telemetry data from 20-25% of
new vehicle sales annually to create a large-scale national sensor dataset. Utilize the
collected data to define safety norms for semi-autonomous vehicles and identify high-
risk roads for infrastructure improvements
• A centrally owned and shared technology stack for Software-Assisted Vehicles (SAVs),
including standardized hardware, software interfaces, and safety regulations to reduce
costs and ensure interoperability, an open SAV reference architecture and a national
high-definition map cloud through public-private partnerships
Talent and Capability Development
• Train and retain a workforce of 30,000+ engineers
33
that specialize in SAVs by 2035
• Centers of Excellence in vehicle software at premier engineering and management
institutes, and introduce SAV-focused minor degree programs
• A Global Mobility Tech Visa to attract international talent with streamlined processing
and competitive compensation
• Upskill and retain engineers with combined expertise in AI and mechanical engineering
by 2035 to support the auto components industry
• Open public-private Centers of Excellence to reskill OEM engineers, integrate AI into
undergraduate curricula, and host industry-supported hackathons
Supplier Ecosystem
• Grow the SAV-ready supplier base to over 100 firms
34
across sensors, ECUs, chip
design, and AI services
• Two auto-grade semiconductor fabs with a monthly production capacity of 40,000+
wafers
35
under the India Semiconductor Mission to meet domestic demand and
potentially create an export surplus
33 Benchmarking against Volkswagen’s 10k engineers for 9M cars and scaling down to 0.6M vehicles/year for an Indian
OEM (i.e., 660 engineers), taking a 20% cut for in-house engineers, yields 100–120 engineers/ OEM for India’s 38-40
OEMs
34 Benchmarking with Germany’s SAV-ready suppliers, scaled to India’s aspiration of producing 5M+ vehicles/year
by 2035; GTAI (Germany Trade & Invest) report titled “Automotive Industry – Germany’s Production of 4.1 Million
Passenger Cars”. October 10 2024.; Meyer Industry Research report titled “TOP 100 Automotive Suppliers Germany
2021”
35 EE Times article titled “ESMC 300-mm Wafer Fab: A Bid to EU’s Semiconductor Sovereignty”. November 7 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 37
Regulatory
• Government decision makers could review standard frameworks for vehicle
cybersecurity and over-the-air software updates to align with global regulations
• Government decision makers could seek to form an SAV Regulatory Taskforce under
the Ministry of Road Transport and Highways to develop agile and globally aligned
standards
• Government decision makers could look at the potential of a fast-track AI patent
regime to reduce the patent grant timeline to under 20 months
Market Access and Compliance
• Concessional GST slab or an income-tax deduction on financing costs for components
designed or validated using AI e.g., Deep-Learning Surrogates (DLS) to encourage
adoption
• “Digital Patent Box” with a lower tax rate for licensed DLS models and AI workflows
developed and commercialized in India
• Global AI safety standards integrated into the existing AIS-140 framework to ensure
compliance and certify AI-generated components for domestic and international
markets
• National Surrogate-Model Certification Sandbox established, e.g., at ARAI Pune, to test
and certify AI models for the automotive industry
Key risks
India’s aspiration of software vehicles is prone to following risks
Data Market access Talent Supplier ecosystem Regulatory and IT Policy
Regulatory misalignment
If the automotive sector does not harmonize
its cybersecurity and software-update rules
with UNECE WP.29 R155/R156 by 2028,
Indian SAVs could fail EU/Japan official
certifications, shrinking exports
36
Domestic supplier base stagnation
If component suppliers do not innovate
quickly enough to develop modular, certified
parts, the penetration of SAVs will remain
limited - slowing roll-outs and constraining
revenue growth
Semiconductor shortfall
If two auto-grade 28/16nm fabs are not
commissioned & global chip procurement
lead time continue to remain high (currently
>70 weeks), assembly lines can idle for 3-5
quarters, reducing value capture
Engineering talent gap
If only 25-30k of the required 35k-40k SAV
engineers are trained or attracted, software
cycles may lengthen 6-12 months, trimming
value capture from subscription revenues
Cyber-trust shock
If a major cyber hack occurs before robust
cybersecurity compliance is in place, it can
crash consumer trust and limit adoption rates.
Feature non-democratization risks
Unless industry bodies actively democratize
and scale Tier 1 and Tier 2 SAV features,
adoption could stay far below the 40–50%
target for 2035
36 UNECE press release titled “Three landmark UN vehicle regulations enter into force”. February 5, 2021 38 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Auto components design
AI-powered models such as Deep-Learning Surrogates (DLS) replicate the behavior of
complex physics simulations, enabling near-instantaneous and highly accurate predictions.
These models are transforming the R&D value chain, redefining how components are designed,
tested, and manufactured. Traditionally, simulating component behavior like aerodynamic
drag, thermal stresses, or structural deformation requires extensive physics-based computing
that can take hours to days per iteration. Once trained, the AI models perform the simulations
in milliseconds, significantly accelerating R&D cycles and enabling more efficient, low-cost
innovation.
For India, which currently accounts for 1-2%
37
of the $500 to $550B
29
global automotive
parts export market i.e., $7-8B, DLS can be an enabler to increase its share significantly. With
automotive parts imports at $6-7B
29
in high-potential areas, there is an opportunity to reduce
import dependency by improving domestic design and testing capabilities. AI-led design,
including DLS, not only boosts competitiveness by slashing development time and costs but
also allows India to lead in high-value, design-driven exports moving beyond assembly and
manufacturing. Indian auto components OEMs can potentially capture $25-30B in cumulative
value by 2035 with an exit value of $4-6B in 2035 (Exhibit 14).
Exhibit 14
37 United Nations Comtrade Database titled “UN Comtrade: International Trade Statistics Database” AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 39
Potential enablers for consideration
Enablers that enhance skill development, strengthen compliance and improve overall
performance could benefit the components sector.
Infrastructure
• Automotive bodies can jointly establish three regional Zero Prototype Labs (ZPLs) in
India’s key component hubs - Chennai/Tamil Nadu (35% revenue share), Pune/Western
belt (33%), and Gurgaon-Manesar (30%). These ZPLs would serve as digital testbeds using
AI, digital twins, and virtual simulations to eliminate early physical prototyping. They can
reduce development time by 20–25% and costs by 10–30%. While large manufacturers
can set up in-house ZPLs, MSMEs and startups will benefit from shared facilities
• Examples of similar lab setups globally include EDAG’s Zero Prototype Lab in Wolfsburg
38
(~1,700 m, 7-9 petaflops of compute power, run 4.5-6K parallel simulations) and Porsche’s
center in Weissach
39
(~2,100 m, 10-12 petaflops of compute power, run 6-8K parallel
simulations)
• The labs can be potentially run on a shared access model with access defined across
three tiers:
pCorporate tier: Paid, high-priority access with dedicated compute and IP security
pMSME & startup tier: Subsidized, affordable simulation slots
pCoE tier: Open academic access for public R&D and skill-building
Incentives and IP Frameworks
• IP for Deep-Learning Surrogate (DLS) models spans data, model architecture, and
digital twin workflows. With significant model training costs, legal protection is key to
driving investment. India could strengthen its incentives and IP protection frameworks
by creating a fast-track AI patent regime to reduce the current patent grant timeline
from 48+ months
40
to less than 18 months, ensuring innovators can secure IP rights
swiftly before data or model designs are leaked or copied
• The shorter timeline would close India’s “protection gap” with the U.S. (about 23 months,
with a 12-month “prioritized” track for digital technologies) and South Korea (standard
16 months)
Compliance
• Physics-Informed & Explainable AI (PI-XAI) means AI models used in vehicles should not
only be as accurate as traditional simulations like Computational Fluid Dynamics (CFD)
but also follow basic physical laws and clearly explain their predictions. For instance,
a 2024 study showed a physics-based AI model could predict battery health with just
0.87% error
41
. But experts highlight that regulators still need clear safety limits and ways
to measure uncertainty before certifying these models.
38 EDAG article titled “Zero Prototype Lab” (part of “Reducing time-to-market”). November 21 2023
39 Porsche Newsroom article titled “Driving simulators: Test drive without a test vehicle”. May 2 2025
40 Times of India report titled “At over 90k patent filings, highest in 2 decades”. January 2024
41 Nature Communications article titled “Physics-informed neural network for lithium-ion battery degradation stable
modeling and prognosis,” Published May 21, 2024 40 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• To encourage compliance, India could integrate global AI safety standards like ISO/PAS
8800:2024 (for explainability and transparency in vehicle AI) and ISO/ IEC 42001 (for
ethical AI management) into its existing AIS-140 framework
42
. This would help suppliers
certify AI-generated components for both Indian and global markets. This is in line
with Germany’s Technischer Über-wachungsverein Süd (TUV SUD) (German technical
inspection and certification organization), which already runs ISO 8800-aligned “AI
Quality” audits for automotive suppliers
43
• Setting up a National Surrogate- Model Certification Sandbox (NSCS) at ARAI Pune
to test and certify AI models using open datasets, physics checks, and cybersecurity
setups, feeding results into the new AIS-XAI approval process.
• The above requirements are not exhaustive; industry organizations could lead the
collation of a complete set of relevant global standards and certifications, and ~3
months under its accelerated route for semiconductor and Artificial Intelligence filings
Talent
• Large Indian auto-component manufacturers can build dual learning tracks, micro-
credentials in DLS for all R&D engineers, and bootcamps for business-unit heads.
This mirrors a global auto components manufacturer’s upskilling of 130K+ employees,
and AI leadership programs at leading automobile OEMs, which have reached 50K+
managers. Indian automotive organizations can establish 20+ public-private Centers of
Excellence (CoEs) hosted in IITs, NITs, IISc, and IIMs/ISB to:
pReskill engineers via semester-long residencies
pEmbed AI–mechanical modules in undergraduate curricula
pRun Formula-Student style hackathons using DLS-only validation, supported by
industry datasets and grants
• Draw inspiration from global automobile OEMs that have collaborated with technical
universities to foster joint research in AI and autonomous systems, or have focused
on research in sensor and AI integration, which helps automotive firms build in-house
expertise in surrogate modelling and AI-driven simulation
42 ISO/PAS 8800:2024 functional-safety standard for AI in road vehicles. Published December 2024; ISO/IEC
42001:2023 AI-Management-System standard. Published December 2023
43 TÜV SÜD AI Quality Certification Program (AIQCP) training programme. (2024) AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 41
Key risks
India’s aspiration to be AI first automotive component manufacturer is prone to risks.
Data Infrastructure Talent Market access Regulatory and IT Policy
Certification risk
If India auto manufacturers doesn’t launch a
certification sandbox and align with global
AI safety norms, auto parts may miss key EU/
US approval windows, causing 1–2 year delays
and risking export contracts.
Infrastructure gap
If the three Zero Prototype Labs aren’t set up by
2029 due to funding gaps, validation will shift
back to physical testing, slowing time-to-market
by 20–30%, potentially eroding value gains.
Supply chain adoption risk
If MSME suppliers can’t access “DLS-as-a-
service” tools, due to high licensing fees, or
limited technical support, overall adoption
may remain low, potentially limiting India’s
potential value capture.
Geopolitical risk
If future WTO or India–EU rules require firms
to disclose training data for safety audits, &
companies see this as a threat to their IP, they
may limit model exports, shrinking India’s global
market access.
AI model integrity risk
If physics-informed quality checks and secure
update systems aren’t in place, even one
major prediction error or cyberattack could
lead to global recalls and a freeze on using
AI-designed parts. 42 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
WAY FORWARD FOR THE INDUSTRY
The full potential footprint of AI on India’s economy is difficult to visualize, and this report
offers a starting point. It outlines two levers with two priority areas each, but the same
structured approach could be extended to other sectors such as logistics, construction, and
retail. Global benchmarks already illustrate the scale of opportunity: the IMF estimates AI
could lift global GDP growth significantly over the next decade
44
; the World Economic Forum
projects that nearly a quarter of all roles worldwide will change within five years due to AI
adoption
45
; and McKinsey research suggests that generative AI alone could contribute more
than any single technological wave in recent history
46
. These underline the importance of
applying an India-specific perspective to additional sectors to identify use-cases, value pools,
and enabling conditions.
Equally, labour transitions would be central to how India adopts AI. International institutions
estimate that around 35-40% of jobs worldwide are exposed to AI, with higher exposure in
advanced economies and meaningful effects across emerging markets
47
. Projections show
that while AI will create many new roles, it will also displace many existing jobs, particularly
in clerical, routine, and low-skill segments. For India, the challenge would be twofold:
preparing a workforce with advanced digital and AI skills to capture new opportunities,
while simultaneously ensuring that those displaced are gainfully employed through reskilling,
redeployment, or absorption into other growth sectors of the economy.
Finally, productivity gains and innovation must match market creation to translate into
growth. India would need to simultaneously deepen domestic demand and secure stronger
participation in global value chains. This will require alignment of industrial and trade policies,
particularly as global rulebooks evolve quickly. For instance, the European Union’s AI Act will
phase in obligations for general-purpose and high-risk AI systems, and new climate-related
trade measures such as carbon border adjustments are set to shape market access conditions
48
.
How India anticipates and responds to global shifts will influence its competitiveness, its
ability to attract investment, and its standing as a credible global partner in the AI economy.
44 IMF working paper titled “The Global Impact of AI: Mind the Gap”. April 2025
45 World Economic Forum press release titled “Future of Jobs Report 2023: Up to a Quarter of Jobs Expected to
Change in Next Five Years”. April 2023
46 McKinsey report titled “The Economic Potential of Generative AI: The Next Productivity Frontier”. June 14, 2023
47 IMF blog post titled “AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity.” January 14, 2024
48 Website titled “EU Artificial Intelligence Act | Up-to-date developments and analyses of the EU AI Act.” 2025 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 43
APPENDIX
Seven Pillars of the IndiaAI Mission
49
Compute
capacity
Establish a federated public infrastructure of high-end GPU clusters
and launch AIRAWAT (AI Research Analysis and Workbench). As of 15th
September 2025, the mission targets deployment of 38k+ GPUs to support
national compute requirements
Innovation
centre
Develop India-specific foundational models and large language models
(LLMs) tailored to Indian languages and domains, advancing IP creation
Datasets
platform
Build anonymized, consent-based, interoperable public datasets to
fuel model training. This platform underpins innovation, application
development, and trusted AI governance
Application
development
initiative
Build AI solutions for agriculture, healthcare, education, and mobility
through targeted pilots and partnerships with industry.
Future skills
Train 50L+ students and professionals through AI curriculum integration,
data labs, and new certification programs in higher education
Startup
financing
Support 1,000+ AI startups through catalytic funding for deep-tech and AI
product innovation
Safe and
trusted AI
Create frameworks and toolkits to ensure explainable, ethical, and privacy-
preserving AI. This includes audit tools and compliance standards.
49 IndiaAI portal (Government of India initiative) 44 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Glossary
AI: Artificial Intelligence: In this report, AI
includes software capabilities such as machine
learning, data analytics and hardware-driven
automation, such as robotics
GDP: Gross Domestic Product: The total value
of all goods and services produced in a country
in a given time
BAU: Business as Usual: A scenario where
current trends continue without major policy
or technology changes
MSME: Micro, Small, and Medium Enterprises:
Businesses that are classified based on
investment size and number of employees
E2E: End-to-End: A complete process or
system that functions seamlessly from start to
finish
AA: Account Aggregator: A system that allows
secure and consent-based sharing of financial
data between institutions
IDMO:
India Data Management Office: A
body responsible for overseeing non-
personal and anonymized data governance
in India
DEPA: Data Empowerment and Protection
Architecture: A framework enabling users
to share personal data securely with
consent
RBI: Reserve Bank of India: India’s central
banking institution that regulates monetary
policy and ensures financial stability
RBIH: RBI Innovation Hub: A unit under RBI
supporting fintech and technology-led
innovations in the financial sector
ReBIT: Reserve Bank Information
Technology Private Limited: An RBI-owned
tech company managing cybersecurity
and IT infrastructure
AML: Anti-Money Laundering: Regulations
and systems designed to prevent illegal
money transactions and funding sources
GPU: Graphics Processing Unit: A high-
performance processor ideal for AI
workloads, simulations, and image
processing
LLM: Large Language Model: An AI model
that understands and generates human-
like text based on large datasets
API: Application Programming Interface:
A set of tools allowing different software
systems to interact and share data
IBA: Indian Bank Association: An industry
body that represents and coordinates
efforts among Indian banks
IT: Information Technology: The use of
systems and computers for processing,
storing, and transmitting information
PLI: Production-Linked Incentive: A
government scheme offering financial
rewards to boost domestic manufacturing
output
OEE: Overall Equipment Effectiveness:
A measure of how well manufacturing
equipment is utilized in terms of time,
speed, and quality
ML: Machine Learning: A form of AI
where systems learn from data to make
predictions or decisions
AGV: Automated Guided Vehicles:
Driverless vehicles used in factories to
transport materials automatically
OEM: Original Equipment Manufacturer:
A company that makes parts or products
used in another company’s end product AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 45
TSN: Time-Sensitive Networking: A
networking protocol that ensures timely
and reliable data delivery for industrial
automation
HPC Cluster: High-Performance Computing
Cluster: A network of powerful computers
used for complex calculations and AI
model training
AICTE: All India Council for Technical
Education: The regulatory authority for
technical and engineering education in
India
Biomarker: A measurable biological
indicator (molecule, gene, image, signal)
that reliably reflects a normal or pathogenic
process or drug response.
Omics is a collective term for large-scale
data layers such as genomics (DNA),
transcriptomics (RNA), proteomics
(proteins), metabolomics (metabolites)
and epigenomics.
Genomics: Study of an organism’s complete
DNA sequence.
Transcriptomics: Analysis of all RNA
transcripts in a cell or tissue
Proteomics: Large-scale study of protein
expression, structure and interaction.
Metabolomics: Comprehensive profiling of
small-molecule metabolites.
Biobank: Organised repository storing
biological samples and associated data for
research.
FAIR principles: Data management
guidelines— Findable, Accessible,
Interoperable, Re-usable.
IBDC: Indian Biological Data Centre—
national life-science data repository.
INSACOG: Indian SARS-CoV-2 Genomics
Consortium— pan-India viral sequencing
network.
ABDM: Ayushman Bharat Digital Mission—
creates India’s national health-data
infrastructure.
DALY (Disability-Adjusted Life Year): A
composite burden-of-disease metric equal
to the sum of Years of Life Lost (YLL) from
premature death and Years Lived with
Disability (YLD); one DALY represents one
healthy year of life lost.
R&D Scientists: Researchers involved
in various stages of drug development
(discovery to approval)
ANRF: Anusandhan National Research
Foundation – India’s R&D funding
body under the Ministry of Science and
Technology
BIRAC: Biotechnology Industry Research
Assistance Council, a Government of India
agency supporting biotech innovation and
startups.
NIH’s “All of US” program: A U.S. research
initiative collecting diverse health data
from over 1 million people to advance
precision medicine.
Institution-initiated trials: Clinical research
studies led and sponsored by academic
or research institutions rather than
pharmaceutical companies.
Operation Warp Speed: A U.S. government
initiative accelerating COVID-19 vaccine
development, manufacturing, and
distribution.
SEBI: Securities and Exchange Board of
India, the securities market regulator,
ensures investor protection and market
transparency.
Weighted tax deductions: Tax incentives 46 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
allowing companies to deduct multiple
of eligible R&D expenses to encourage
innovation.
Ramanujan Fellowship: Scheme offering
re-entry support to Indian scientists abroad
for research in India
Ramalingaswami Fellowship: Fellowship
program targeting postdocs in life sciences
returning to India
NAMTECH: Finishing-school style institute
model focused on deep-tech and job-
ready skills
AICTE: Apex regulatory body for technical
education in India
OCI-Fellow: Proposed track offering
incentives for Overseas Citizens of India to
return
HPC Cluster: High-performance computing
facility used for large-scale simulation and
AI research
Indian Biological Data Centre (IBDC):
India’s central omics repository, housing
genomics and bioinformatics datasets
PPP-adjusted: Purchasing Power
Parity–adjusted salary comparisons to
international levels
ICH: International Council for
Harmonisation of Technical Requirements
for Pharmaceuticals for Human Use – sets
global drug development guidelines
E2E: ICH guideline on Pharmacovigilance
Planning – ensures safety monitoring
throughout the drug lifecycle
E9: ICH guideline on Statistical Principles
for Clinical Trials – ensures consistency and
quality in trial design and data analysis
E17: ICH guideline on Multi-Regional
Clinical Trials – provides a framework for
conducting global trials efficiently
M2/M8: ICH guidelines for Electronic
Standards & Submission Format – ensures
standardized digital submissions to
regulatory authorities
CDSCO: Central Drugs Standard Control
Organization – India’s national drug
regulatory body
GCP: Good Clinical Practice – international
standard for ethical and scientific quality
in clinical trials
Data Exclusivity: A legal right preventing
generic manufacturers from referencing
an innovator’s clinical data for a defined
period
Hatch-Waxman Act: U.S. legislation that
grants data exclusivity of 5–12 years for
new drugs, protecting clinical trial data
ANRF: Anusandhan National Research
Foundation – India’s upcoming R&D
funding agency to support deep science
initiatives
CDSCO NDCT Rules (2019): India’s rules for
approval of new drugs and clinical trials;
enables accelerated pathways and ethics
compliance
DPIIT: Department for Promotion of
Industry and Internal Trade – involved in IP
reforms and innovation policy
Market Access: A Set of policies and
incentives that determine how quickly
and widely a drug reaches patients at an
affordable price
HTAIn: Health Technology Assessment
India – evaluates the value-for-money of
health technologies
HTA: Health Technology Assessment AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 47
– compares clinical efficacy, cost-
effectiveness, and equity of medical
technologies
DHR: Department of Health Research –
nodal body for HTAIn under Ministry of
Health
PM-JAY: Pradhan Mantri Jan Arogya Yojana
– India’s public health insurance scheme
covering 550M people
CDSCO: Central Drugs Standard Control
Organization – India’s national regulatory
body for drug approvals
NHA: National Health Authority –
implementing agency of PM-JAY
NRDL (China): National Reimbursement
Drug List – HTA-based system for national
drug price negotiation and reimbursement
NICE (UK): National Institute for Health
and Care Excellence – UK’s HTA body with
legal timelines post EMA approval
EMA: European Medicines Agency – central
drug approval authority for the EU
NITI Aayog: Policy think tank of the
Government of India – drives long-term
policy vision
Made-in-India Innovation Track: Proposed
track for public coverage of Indian-
developed drugs within 180 days post
CDSCO approval
DPIIT: Department for Promotion of
Industry and Internal Trade – responsible for
Startup India and pharma manufacturing
schemes
PLI: Production Linked Incentive – scheme
to boost local manufacturing with
performance-based incentives
Proprietary AI model: An AI system
exclusively owned and trained by an entity,
optimized for a specific application like
drug discovery
Foundation model: A large-scale pre-
trained AI model that can be adapted to
multiple downstream tasks
Genome-scale language model: An AI
model trained on DNA/RNA sequences to
predict the functional impact of mutations
and guide target discovery
Nucleotide Transformer: A 2.5B parameter
AI model developed by InstaDeep, NVIDIA,
and TUM that sets a new benchmark in
variant effect prediction
Single-cell foundation model: A model
trained on individual cell data (multi-omics)
to predict how drugs act across different
cell types
scGPT: A “ChatGPT for biology” AI model
developed by the Toronto Vector Institute
and UHN, trained on millions of single-cell
data points
Protein-design model: AI model that
predicts 3D protein structures and
interactions to help design new drugs
AlphaFold 3: DeepMind’s AI model that
accurately predicts 3D structures and
interactions of proteins and small molecules
Multimodal clinical-prediction model: An
AI system that integrates omics, EHR,
and phenotype data to guide clinical trial
inclusion and adverse event forecasting
Bridge2AI: U.S. NIH initiative to develop
AI-ready health datasets for training
multimodal biomedical foundation models
India Omics Commons Sandbox: Proposed
national platform to house diverse omics
datasets for AI model development
SAV (Software Assisted Vehicle): A vehicle
with embedded software capabilities, 48 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
including smart features like predictive
maintenance, OTA updates, and data
services
OTA (Over-the-Air): Wireless delivery of
software updates or patches to a vehicle’s
system without requiring service center
visits
Modular Components: Vehicle parts
designed to be independently upgraded,
replaced, or certified
Certified Components: Parts that
meet standardized safety, quality, and
performance benchmarks set by regulatory
bodies
ECU (Electronic Control Unit): Embedded
systems in vehicles that control electrical
subsystems like braking, engine, and
infotainment
ISM (India Semiconductor Mission):
A government initiative to build a
comprehensive semiconductor ecosystem
in India
28nm / 16nm: Semiconductor process
nodes indicating the transistor size used in
chip fabrication; 28nm/16nm are common
for automotive-grade chips
Wafers per Month (WPM): Unit of fab
output; indicates how many silicon wafers
can be produced each month
Brownfield Fab: A semiconductor plant
built on an existing industrial site, often
through upgrades or repurposing
Greenfield Fab: A new chip manufacturing
facility built from scratch
CHIPS and Science Act: A U.S. federal law
enacted in 2022 that allocates $52 billion
to promote domestic semiconductor
manufacturing
SMIC (Semiconductor Manufacturing
International Corporation): China’s largest
chipmaker, known for strategic investment
in 28nm and 40nm auto chips
ACMA: Automotive Component
Manufacturers Association of India; apex
body for India’s auto component suppliers
SIAM: Society of Indian Automobile
Manufacturers; works on policy and
technical issues affecting Indian OEMs
ARAI: Automotive Research Association
of India; certifies automotive components
and systems
BIS: Bureau of Indian Standards; national
standards body responsible for setting
quality norms across sectors
ISMC: International Semiconductor
Consortium, a public-private partnership
aiming to establish fabs in India
SAV: Software-Assisted Vehicle; vehicles
integrated with software-based features
like infotainment, diagnostics, and driver
assistance “Intelligent and Connected
Vehicles
Cyber-Physical System (ICV CPS):
Vehicles embedded with AI, sensors,
and communication tech to interact with
infrastructure and other vehicles in real
time.”
NIC: National Informatics Centre, India’s
technology backbone for e-governance
and digital infrastructure across
government bodies.
ISRO: Indian Space Research Organisation,
India’s national space agency responsible
for space missions, satellites, and launch
vehicles.
Zero prototype lab: A digital R&D setup
where designs are validated entirely
through simulations and AI without AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 49
building physical prototypes.
Tier 1/2 Suppliers: Automotive component
manufacturers directly (Tier 1) or indirectly
(Tier 2) supplying parts to OEMs
Validation Cycle: One complete test cycle
of a SAV feature (e.g., parking assist) under
various simulated conditions
RSU (Roadside Unit): Fixed communication
infrastructure that connects vehicles to
network and surroundings for real-time
data exchange
5G/6G Corridors: High-speed connectivity-
enabled roads to support real-time
software and V2X (vehicle- to-everything)
communication
NATRAX: National Automotive Test Tracks,
a state-of- the-art test facility in Indore,
India
STPI: Software Technology Parks of India;
organization supporting IT/ITES industries
with infrastructure and services
American Center for Mobility (ACM): A 500-
acre testing site in Michigan, USA, focused
on software- defined vehicle technologies
K-City: An 80-acre autonomous driving
testbed in South Korea, equipped for
smart mobility trials
DLS (Deep Learning Surrogates): AI
models that emulate physical simulations
to accelerate design and validation of
components
OEM (Original Equipment Manufacturer):
A company that produces parts and
equipment that may be marketed by
another manufacturer
CoE (Centre of Excellence): Specialized
institutional setups for applied research,
talent development, and innovation
AICTE: All India Council for Technical
Education – a body for curriculum reforms
and engineering standards
Micro-credentials: Short, specialized
certifications validating skillsets in niche
areas like AI in engineering
Hackathon: A competitive, time-bound
innovation challenge to solve real-world
problems
Formula-Student: An international student
competition to design and race small-
scale formula-style cars
Digital twin: A virtual representation
of a physical object or system used for
simulation, design testing, and monitoring
High-performance computing (HPC):
Large-scale computational infrastructure
used to train complex AI models and run
massive simulations
Patent grant timeline: The time taken from
patent filing to final approval and grant of
IP rights
AI-priority patent track: A fast-track
mechanism for processing patent
applications related to artificial intelligence
innovations
Model licensing scheme: A framework
enabling OEMs or Tier-1 suppliers to license
trained AI models or virtual component
validations as services
Intellectual property (IP): Legal rights
granted over inventions, datasets, models,
and processes that are original and provide
economic benefit
Ministry of Electronics and Information
Technology (MeitY): Indian ministry
overseeing digital and AI technology
advancement
Department for Promotion of Industry and 50 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Internal Trade (DPIIT): Indian ministry
responsible for IP frameworks, industry
regulations, and startup innovation policies
PI-XAI: Physics-Informed & Explainable AI
— AI models that both obey physical laws
and provide interpretable results
CFD: Computational Fluid Dynamics —
Traditional physics-based method used
to simulate fluid flow and heat transfer in
automotive and other systems
ISO/PAS 8800:2024: International standard
for transparent and explainable AI systems
used in vehicles
ISO/IEC 42001: International standard for
ethical and accountable AI management
within organizations
AIS-140: Indian Automotive Industry
Standard related to Intelligent Transport
Systems, including GPS tracking and
emergency systems
AIS-XAI: Proposed extension of AIS
standards to include explainable AI
certifications
TÜV SÜD (Germany): Technical inspection
and certification agency offering AI
quality audits compliant with international
standards
NSCS: National Surrogate-model
Certification Sandbox — A proposed
setup at ARAI for testing and certifying AI
models using open data, physics checks,
and cybersecurity
BIS: Bureau of Indian Standards — India’s
national standards body responsible for
standard setting and certification
MHI: Ministry of Heavy Industries — Central
ministry responsible for the automotive
industry’s policy and standards governance
GST (Goods and Services Tax): A unified
indirect tax in India levied on the supply of
goods and services
Digital Patent Box: A policy regime offering
lower tax rates on profits earned from
locally developed and licensed IP
CBDT: Central Board of Direct Taxes,
responsible for direct tax policy formulation
and enforcement
PLI (Production Linked Incentive): A
government scheme offering financial
incentives based on incremental
production and sales
ICE (Internal Combustion Engine):
Traditional vehicle engines that run on
petrol or diesel
Bureau of Indian Standards (BIS): National
standards body responsible for product
certification schemes
NITI Aayog: India’s apex policy think tank
guiding economic and technological
policy reforms
Ministry of Heavy Industries (MHI): Nodal
ministry for the automotive sector, among
others
Digital twin: A virtual representation
of a physical object or system used for
simulation, design testing, and monitoring
Patent grant timeline: The time taken from
patent filing to final approval and grant of
IP rights
AI-priority patent track: A fast-track
mechanism for processing patent
applications related to artificial intelligence
innovations
Model licensing scheme: A framework
enabling OEMs or Tier-1 suppliers to license
trained AI models or virtual component
validations as services AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 51
Intellectual property (IP): Legal rights
granted over inventions, datasets, models,
and processes that are original and provide
economic benefit
Industry 4.0: The fourth industrial revolution
is marked by integration of digital
technologies like AI, IoT, and robotics into
manufacturing and industry.
Advanced Cyberinfrastructure Coordination
Ecosystem-Services & Support (ACCESS):
A U.S. initiative providing advanced
computing, data, and support services to
accelerate scientific research.
NEAT: National Educational Alliance
for Technology, a Government of India
platform to bring AI-powered learning
tools to higher education 52 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
NOTES AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 53
NOTES 54 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
NOTES AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 55 56 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
W
e are indebted to the Expert Council for its strategic
foresight and instrumental contributions in shaping this
project. Their guidance has ensured that the roadmap
reflects both ambition and pragmatism, making the recommendations
highly actionable.
We also gratefully acknowledge the invaluable inputs of the
following domain experts: Dr Bala Subrahmanya, General Manager,
ABLE; Mr GS Krishnan, President, ABLE; Ms Kiran Mazumdar-Shaw,
Executive Chairperson, Biocon Ltd; Mr Mandar S Ghatnekar, Global
Head of IT & Digital Transformation, Biocon Biologics; Mr Shailendra
Trivedi Chief General Manager-in-Charge Department of Information
Technology, RBI and Mr Suvendu Pati Chief General Manager
Financial Technology Department, RBI; whose insights enriched the
work and firmly anchored it in ground realities.
A special word of thanks is extended to McKinsey & Company for its
exemplary partnership in providing analysis & insights that helped
us to develop a roadmap that is implementable, impact-driven, and
aligned with India’s long-term vision.
Together, the collective wisdom and collaboration of all partners
have made this effort possible. 4 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth Mr. N Chandrasekaran
Chairman, Tata Sons
Mr. Neelkanth Mishra
Chief Economist, Axis Bank and Head
of Global Research, Axis Capital
Mr. Rahul Matthan
Partner, Trilegal
Dr. Anish Shah
Managing Director and CEO,
Mahindra & Mahindra Ltd
Ms. Anu Madgavkar
Partner, McKinsey Global
Institute
Mr. Noshir Kaka
Senior Partner, McKinsey &
Company
Mr. Ishtiyaque Ahmed
Program Director, Industry/
MSME, NITI Aayog
Mr. Chandrajit Banerjee
Director General, CII
Mr. Mukesh Bansal
Founder - Myntra &
CureFit
Dr. Chintan Vaishnav
Former Mission Director, Atal
Innovation Mission, NITI Aayog
Expert Council Members I
f India is to accelerate its growth to the 8% annual rate
required for the realization of Viksit Bharat, we have
no option but to significantly raise productivity across
the economy and unlock new growth through innovation.
Artificial Intelligence can be the decisive lever. This report
sets out a practical roadmap on how we can harness Al to
translate this potential into outcomes.
The analysis highlights two major Al unlocks. First,
accelerating adoption of Al across industries to enhance
productivity and efficiency-bridging nearly 30-35% of the
required step-up. Second, transforming R&D, especially
through generative Al, which can enable India to leapfrog
into innovation-driven global opportunities, contributing
at least 20-30% of the required uplift.
With a focused and sector-specific approach, industries
such as banking and manufacturing can deploy Al today to
improve efficiency, service quality, and competitiveness-
creating momentum for deeper transformation. At the
same time, India must nurture frontier innovation, from
Al-enabled drug discovery to software-defined vehicles,
building the next engines of growth.
The path to 8% growth runs through decisive Al adoption
and innovation. This report offers a roadmap to guide that
journey. I invite government, industry, and academia to
move forward with urgency and collective purpose.
BVR Subrahmanyam
CEO, NITI Aayog
Foreword T
he NITI Frontier Tech Hub’s AI roadmap for Viksit
Bharat sends an unequivocal signal: India’s mission
to sustained 8%+ growth is anchored in bold,
pervasive AI integration and tireless innovation—and
must become a core national priority. This transformation
journey leverages sector-focused strategies and frontier
technology ecosystems, positioning India to lead the
global race in inclusive, responsible AI deployment and
governance.
Our deep gratitude is due to the Expert Council for its
strategic foresight and instrumental contributions, and
to McKinsey for its exemplary partnership in shaping a
roadmap that is implementable and impact-driven.
The time for India to lead the AI revolution at scale is now.
With robust policy frameworks, advanced infrastructure,
and collaborative innovation, India can pioneer a new
model of growth and societal advancement, ensuring
prosperity, resilience, and technological leadership for
decades to come.
The NITI Frontier Tech Hub will continue to activate this
agenda, galvanizing experts, states, and industry toward
shared progress—securing the foundations for an AI-
powered Viksit Bharat
Debjani Ghosh
Distinguished Fellow, NITI Aayog;
Chief Architect, NITI Frontier Tech Hub
Foreword Index
CHAPTER 1: INTRODUCTION....................................................................................10
AI OPPORTUNITIES FOR INDIA..........................................................................................10
STRATEGIC ENABLERS FOR AI-LED VALUE CREATION.........................................11
CHAPTER 2: POTENTIAL OUTCOMES FOR AI-LED VALUE CREATION............12
POTENTIAL OUTCOME 1: INDIA BECOMES
THE DATA CAPITAL OF THE WORLD..............................................................................13
POTENTIAL OUTCOME 2: INDIA SUPPORTS THE DEVELOPMENT OF AN
ADAPTABLE AND EFFICIENT AI-SKILLING ECOSYSTEM.......................................14
POTENTIAL OUTCOME 3: TARGETED AI ADOPTION UNLOCKS SECTORAL
GROWTH......................................................................................................................................14
POTENTIAL OUTCOME 4: INDIA’S JOBS ARE FUTURE-PROOFED, AND
INDUSTRY TRANSFORMED AT SCALE...........................................................................15
CHAPTER 3: LEVER 1 - ACCELERATING AI ADOPTION
ACROSS INDUSTRIES TO IMPROVE PRODUCTIVITY AND EFFICIENCY..........17
BANKING.....................................................................................................................................19
POTENTIAL ENABLERS FOR CONSIDERATION.....................................................21
KEY RISKS..............................................................................................................................23
MANUFACTURING....................................................................................................................24
POTENTIAL ENABLERS FOR CONSIDERATION.....................................................25
KEY RISKS..............................................................................................................................26 CHAPTER 4 – LEVER 2: UNLOCK LEAPFROG INNOVATION
BY TRANSFORMING R&D WITH AI..........................................................................27
PHARMACEUTICALS..................................................................................................................29
POTENTIAL ENABLERS FOR CONSIDERATION........................................................31
KEY RISKS. ................................................................................................................................32
AUTOMOTIVE................................................................................................................................32
SOFTWARE-ASSISTED VEHICLES (SAVS)................................................................... 33
POTENTIAL ENABLERS FOR CONSIDERATION..................................................35
KEY RISKS............................................................................................................................37
AUTO COMPONENTS DESIGN..........................................................................................38
POTENTIAL ENABLERS FOR CONSIDERATION..................................................38
KEY RISKS............................................................................................................................41
WAY FORWARD FOR THE INDUSTRY................................................................... 41
APPENDIX...................................................................................................................43 10 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
CHAPTER 1: INTRODUCTION
Over the next decade, the adoption of Artificial Intelligence (AI) across sectors is expected
to add $17–26T
1
to the global economy. India’s combination of a large STEM workforce,
expanding R&D ecosystem, and growing digital and technology capabilities positions the
country to participate in this transformation, with the potential to capture 10–15% of global
AI value.
1
Placed against India’s economic outlook, this potential becomes more significant. At its
current growth rate of 5.7%, India’s GDP is projected to reach $6.6T by 2035. However,
under the aspirational 8% growth trajectory outlined in the government’s vision for the nation
known as Viksit Bharat, India’s GDP could increase to $8.3T, representing an incremental
$1.7T compared with the current growth path (Exhibit 1)
AI opportunities for India
Potential AI opportunities for India are presently spread across three levers:
1. Accelerating AI adoption across industries to improve productivity and efficiency,
potentially bridging 30–35% of the gap: Higher output, lower costs of goods and
services, and improved access for underserved markets. These effects are expected to
materialize across both domestic consumption and export markets
2. Transforming R&D, through generative AI, could help India leapfrog into innovation-
driven global opportunities, bridging a minimum 20–30% of the gap: Can generate new
AI-led market opportunities within traditional industries, support commercialization,
reshape legacy value chains, and strengthen long-term competitiveness
3. Innovation in technology services, strengthening India’s reputation as a technology
services leader, contributing another 15-20% to the step up: Could drive the
development of higher-value solutions and new business models, enhancing India’s
competitiveness in the global market
This roadmap focuses on the first two levers, while a separate publication addresses the
third—innovation in technology services.
This roadmap is the first version of this perspective. The insights and recommendations in this
report will be periodically revised, to reflect evolution of the technology as well as the global
economic context. This will keep India’s strategy for accelerated economic development
relevant, resilient and future-ready.
1 McKinsey report titled “The Economic Potential of Generative AI: The Next Productivity Frontier”. June 2023. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 11
Exhibit 1
Exhibit 2
Strategic enablers for AI-led value creation
Realizing the potential of AI in India depends on establishing strategic enablers across
infrastructure, governance, industry, and workforce development. Effective collaboration
between government, the private sector, and academia can support responsible deployment, 12 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
scaling, and skill development, while ensuring broad access and alignment with national
priorities.
• Access to critical AI infrastructure such as cloud platforms, compute, and foundational
datasets could help strengthen India’s sovereign AI capabilities. At the same time,
robust AI governance frameworks, including ethics guidelines and risk controls, could
ensure responsible and secure deployment.
• The private sector can lead the scaling of AI adoption by embedding AI into core
industry processes. This includes driving model validation, secure deployment, and AI-
powered decision-making, while maintaining resilience and accountability at leadership
levels. Reskilling senior executives and upskilling the broader workforce would be key
to enabling this transformation.
• Academia can be vital in anchoring research and supporting large-scale workforce
transformation. The creation of AI testing sandboxes can further enable safety and
scale.
• To ensure inclusive growth, it is essential to provide equitable access to AI resources
and opportunities, particularly for MSMEs and economically underrepresented regions.
These enablers can be potentially mapped into a phased possible path forward, covering
short-, medium-, and long-term priorities aligned with India’s 2035 goals. Progress could be
tracked against established KPIs, with relevant baselines.
India is at a pivotal point in its AI journey. It can capture a meaningful share of AI-driven value
by leveraging its strengths and implementing key enablers. The following chapters explore
the industries, business models, and approaches that could support this transformation.
CHAPTER 2: POTENTIAL OUTCOMES FOR AI-LED VALUE CREATION
AI remains in its formative stage, and market structures are still evolving. To secure a leading
position globally, India could consider investing in sovereign infrastructure, including energy,
to build resilience and unlock higher value-creation potential.
The India AI Mission, with an estimated budget of over ₹10,000 Cr for five years, represents
a foundational step toward strengthening national AI capabilities by focusing on data, talent,
and adoption. Built on seven core pillars, (as of 15th September 2025) the mission plans to
deploy 38,000+ GPUs
2
through a federated compute network, develop India-specific large
language models, and establish an anonymized, consent-based public dataset platform,
placing data at the center as a key enabler for innovation, scalability, and governance in
a diverse, multilingual nation. The initiative also aims to expand AI skilling through the
integration of AI courses at multiple academic levels and creation of AI/Data Labs across Tier
2 and Tier 3 cities, while accelerating adoption through application development in critical
sectors such as agriculture, healthcare, education, and mobility.
2 Press Information Bureau (PIB), Government of India. “Cabinet approves IndiaAI Mission – a significant step towards
boosting India’s AI ecosystem.” March 7, 2024. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 13
While these measures would help lay the groundwork for robust infrastructure, it is also
important that talent pipelines, sectoral uptake, sustained execution, and ecosystem
alignment are in place to realize India’s AI potential. In the following section, the report
identifies potential outcomes that could contribute to the AI-led GDP growth by 2035.
Potential Outcome 1: India becomes the data capital of the world
One of India’s biggest strengths is its data, and it has the potential to leverage this advantage. In
the digital economy, data will function as currency that powers innovation, drives valuations,
and shapes global leadership. India has the potential to lead due to its scale, diversity and
digital infrastructure. By placing quality, trusted, and interoperable data at the core, India
could become the data capital of the world and set new global benchmarks for breadth,
depth, and quality of trusted data ecosystems, potentially through the following options:
• Creating an anonymized data collection framework to easily and safely collect public
data led by entities such as India Data Management Office (IDMO) and National Data
Access Platform
• Building a marketplace of certified non-personalized data with privacy tags and quality
certificate features supported by the National Data Governance Framework
• Developing specialized data platforms for specific sectors as follows:
pFinancial services: Enabling access to cross-industry, alternative data sources with
the IDMO standardizing access and ensuring borrowers retain the right to opt out
pManufacturing: Establishing an open “Manufacturing Data Grid” for OEMs, suppliers,
and startups to trade production and supply-chain data through standard APIs
pPharmaceuticals: Building a unified national omics dataset by sequencing over
10M genomes by 2035 to fuel AI in drug discovery
pAutomotive: Enabling OEMs to share anonymized telemetry data to create a
large-scale sensor dataset for safety and innovation
As of 15th September 2025, AI Kosh, owned and operated by the Government of India under
the India AI Mission, hosts over 2000 curated, non-personal datasets such as census data,
Indian language resources, and satellite imagery.
3
While this is a good foundation, scaling
the breadth and depth of data could position it to move into high-value domains such as
genomics, manufacturing telemetry, and cross-sector financial data, implying that datasets
are certified for quality, tagged for privacy, and interoperable. This can potentially transform
AI Kosh from a foundational national repository into a most trusted and innovation-ready
data platform. India could consider the following options:
• Setting up sector-specific data infrastructure (e.g., for the financial sector) that is
integrated with AI Kosh to host regulatory-grade datasets for model building and
supervision.
• Publishing institution-level AI inventories and publishing a sector-wide AI repository
3 IndiaAI Mission article titled “Now Open: Expression of Interest (EOI) to Contribute Datasets and AI-Artefacts to
AIKosh”. July 7, 2025 14 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
(metadata only) to provide supervisory visibility as AI scales across sectors.
• Integrating AI with Digital Public Infrastructure (DPI) to accelerate inclusive, affordable
financial services at scale (e.g., voice access, Account Aggregator flows, fraud controls).
Potential Outcome 2: India supports the development of an adaptable and efficient
AI-skilling ecosystem
By 2035, India can work towards narrowing the AI skill gap with leading countries by
developing skilled professionals, advancing research, and contributing to AI models. The
focus can be on measuring outcomes of research impact basis international peer-reviewed
publications, the number of PhDs focused on AI research, practical expertise and original AI
contributions based on patents filed. India could consider the following options:
• Harnessing academia better: AI Chairs across the top 20 technology, medical, law, and
business schools to promote and create PhDs/senior qualifications at the intersection
of subject matter expertise and AI
• Incentivizing industry: Finding ways to upskill 3–5% of working professionals through
AI-first modules, supported by tax deductions on employer spending
• Upskilling India: An AI Open University (both physical and online) could enhance Skill
India Digital/eShram in partnership with private ed-tech platforms for the public
• Equipping professionals: Specialized skilling across sectors, illustrated below through
two key examples:
pFinancial services: Launch a national certification program in AI for Credit, Risk
and Fraud, co-designed with top universities
pManufacturing: Initiate a tiered “AI for Advanced Manufacturing” credential in
collaboration with industry leaders
• Exploring AI governance literacy: Look at how to educate Boards and C‑suite of
regulated financial entities covering accountability, risk management, deployment
models (e.g., human-in-the-loop), documentation, and disclosures.
• Ensuring oversight: Consider how a new supervisor entity could help to build capacity
in AI oversight, model risk, audits, and sector risk intelligence.
Potential Outcome 3: Targeted AI adoption unlocks sectoral growth
Focusing on AI enablers across the manufacturing, financial services, pharmaceuticals, and
automotive industries—which represent roughly 25% of India’s projected 2035 GDP—can
help translate AI adoption into measurable outcomes by supporting innovation, improving
productivity, and enhancing export potential. A sectoral analysis highlights these areas as
potential candidates for AI-led innovation and accelerated growth (Exhibit 4).
• Manufacturing: Building world-class smart-factory corridors could enable AI growth
unlocks by: (1) Designating AI-ready industrial parks that co-locate clean-energy
plants with high-performance-computing labs and robotics test zones; (2) Launching
an open manufacturing data grid so OEMs, suppliers and start-ups can exchange real-
time production data via standard APIs; and (3) Rolling out a tiered “AI for advanced AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 15
manufacturing” credential to up-skill engineers from micro-learning to postgraduate
level.
• Financial services: Unlocking responsible AI at scale by: (1) Scaling-up sandbox pilots
to pressure-test explainable credit, fraud-risk and anti-money-laundering models; (2)
Introducing or enhancing existing frameworks to share alternative, consent-based
datasets; and (3) Certifying specialists through a national “AI for Financial Services”
program, creating a trusted talent bench for financial institutions.
• Pharmaceuticals: AI could unlock growth by compressing drug-discovery timelines
and investments by: (1) Expanding biotech parks by 10x and adding best-in-class,
high-performance computing to power AI modelling; (2) Creating a unified national
omics dataset with tiered researcher access; and (3) Training over 100k biopharma
R&D scientists in computational biology and AI by 2035.
• Automotive: Leapfrog to software-led, autonomous mobility by: (1) Setting up six to
eight physical-digital testing parks for India-specific autonomous-vehicle validation;
(2) Deploying 10,000 km of 5G/early-6G “smart corridors” for real-time vehicle-to-
infrastructure data flow; and (3) Mandating anonymized telemetry from 20–25% of
new vehicles each year to seed national safety and innovation datasets.
This analysis could be applied to construction, wholesale and retail trade, and professional
services, contributing another 25% of the projected 2035 GDP and ensuring AI-driven value
scales across the economy (Exhibit 4).
Potential Outcome 4: India’s jobs are future-proofed, and industry transformed at scale
India can address fragmented skilling through a unified system that enables continuous
worker upskilling, accelerates firm-level digital adoption, and strengthens safety nets. This
would mean mapping job shifts annually, embedding lifelong learning into career pathways,
scaling MSME digital upskilling, and protecting gig and platform workers – projected to reach
about 23.5M by 2029–30
4
. International benchmarks highlight both the urgency and the
opportunity: the World Economic Forum estimates 23–25% of roles will change within five
years (69M created; 83M disrupted)
5
, the OECD finds 27% of jobs are in occupations at high
risk of automation
6
, the U.S. may see 12M occupational transitions by 2030 as generative
AI scales
7
, and earlier global studies suggest China could face up to ~100M such transitions
under fast automation scenarios
8
. India could consider the following options:
4 NITI Aayog policy brief titled “India’s Booming Gig and Platform Economy: Perspectives and Recommendations on
the Future of Work”. June 2022
5 Future of Jobs Report 2023 titled “Future of Jobs Report 2023: Up to a Quarter of Jobs Expected to Change in
Next Five Years”. April 30, 2023.
6 OECD report titled “OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market”, finding that
approximately 27% of jobs across OECD countries are in occupations at high risk of automation (including AI). July
2023
7 McKinsey Global Institute report titled “Generative AI and the future of work in America”. July 26, 2023.
8 CSIS ChinaPower Project analysis titled “Is China Ready for Intelligent Automation?” August 25, 2020 16 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• Continuous reskilling: Developing job transformation maps for 25-30 priority sectors
can help to identify task shifts, emerging roles, and reskilling pathways, drawing on
models such as “Workforce Singapore”
9
. Large enterprises could prepare and submit
firm-level skilling plans.
• Accessible learning: One approach could be introducing digital, portable individual
learning accounts. Singapore’s “SkillsFuture”
10
and the UK’s “Lifelong Learning
Entitlement”
11
offer interesting models to consider. These accounts could be designed
to link credit reimbursements to verified course completion and employment outcomes.
• Industry-wide AI adoption: By preparing industry AI plans, with curated toolkits, vetted
vendor lists, and sector-specific micro-credentials. Consider Singapore’s “Industry
Transformation Map” framework as a reference
12
.
• Supporting gig and platform workers: By implementing the Code on Social Security
(2020)
13
, ensuring universal registration and benefits such as health insurance, accident
cover, and retirement savings that can be carried between jobs. Consider using the
e-Shram portal to manage benefits and pilot wage-loss insurance schemes for workers
facing job loss.
• Supporting at-risk worker groups, a live, integrated skills and jobs database by
linking e-Shram, Skill India Digital, and public job listings into a constantly updated
system that tracks in-demand skills, identifies at-risk worker groups, and connects
them directly to funded training and verified job opportunities.
• Financial consumer and worker AI literacy. Ensuring disclosures and grievance
pathways into adoption programs so that trust leads to usage.
Guided by these potential outcomes, the report proceeds to examine two primary AI
opportunity levers for India: Lever 1 focuses on accelerating AI adoption across industries
to enhance productivity and efficiency, while Lever 2 explores the transformation of R&D
through generative AI, enabling India to capture innovation-driven opportunities and bridge
a significant portion of the growth gap.9 Workforce Singapore website
10 SkillsFuture Singapore (“SSG”) government website
11 House of Commons Library research briefing titled “The Lifelong Learning Entitlement.” Published 12 March 2024.
12 Singapore Ministry of Trade and Industry webpage titled “Overview” (Industry Transformation Maps under the S$4.5
billion Industry Transformation Programme). Last updated September 8, 2025
13 Gazette of India (Extraordinary), titled “The Code on Social Security, 2020 (No. 36 of 2020)”. 28 September 2020. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 17
CHAPTER 3: LEVER 1 - ACCELERATING AI ADOPTION ACROSS INDUSTRIES
TO IMPROVE PRODUCTIVITY AND EFFICIENCY
To assess AI’s potential for India, a detailed analysis was conducted on its ability to enhance
productivity across industries. The study covered over 850 occupations across 16 sectors and
examined more than 2,100 distinct work activities. The analysis indicates that AI adoption
could contribute an additional $500–600B to India’s GDP by 2035, beyond the projected
growth trajectory, driven by productivity improvements, operational efficiencies, and the
reallocation of human effort to higher-value tasks.
The following sections detail the methodology and findings behind these projections.
Approach
Specific adoption scenario models were considered (i.e., the pace at which industry adopts
the technology at scale, resulting in impact on productivity)—early, midpoint, and late—to
estimate when AI could effectively take on these activities based on currently demonstrated
technologies and their expected development in the future and country-specific factors such
as wage levels and occupational mix. The model incorporates software capabilities such as
machine learning, data analytics and hardware-driven automation such as robotics.
• Baseline employment and GDP data: Used 2022 as the baseline year for both
employment and real GDP, sourced from IHS. Calculated productivity as GDP per
worker to set the reference point for future projections
• AI adoption rates for 2035: Estimated sector-level AI adoption rates using McKinsey
Global Institute’s (MGI) model, covering about 850 occupations and 2,100 activities
with sectoral nuances. Applied these AI adoption rates to baseline employment to
determine workforce segments likely to be automated
• Growth rates across different scenarios: Augmented workforce calculated by
redeploying the automated workforce at current productivity levels. New GDP projected
by applying 2022 productivity to the augmented workforce. This yielded GDP CAGR
over 2022–2035, forming the basis of Lever 1 sectoral projections
• Relevant scenarios chosen (illustrated below in Exhibit 3): For each sector, AI
adoption across early, mid, and late horizons, under two cases were modeled- AI
adoption scenarios across late, mid and early for each sector. Acceleration is assumed
at two levels, leading to two scenarios: the accelerated AI adoption scenario assumes
faster tech adoption with sectors shifting to earlier phases by 2035, while the moderate
scenario assumes slower adoption and later starting points
• GDP 2024 and business-as-usual CAGR 2024-2035: Estimated sector-wise data from
the IHS database, extrapolated to align with government projections for a total of
$3.6T with sectoral nuances
• Final incremental AI productivity impact value: Derived using an additional productivity
boost to the expected business-as-usual CAGR and then applying it to the GDP 2024
numbers to receive GDP with AI adoption 18 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Assumptions
Below are the key assumptions that have been considered in the model (Exhibit 3):
• The model evaluates current capabilities required to automate tasks in each sector
based on their complexity and AI readiness
• Technologies are assessed on whether they can match or exceed the performance of
top-quartile human workers for specific tasks
• Adoption considers the estimated time to build viable AI solutions and whether they
are economically justified from a cost-benefit perspective
• The degree of AI adoption in a sector is closely linked to its occupational structure (i.e.,
task types and automatable roles)
• Displaced workforce due to AI is assumed to be redeployed within the same sector,
maintaining existing sectoral labor productivity levels
• While India has historically seen slow AI adoption, sector-specific improvements in
digital infrastructure and technology maturity can enable faster adoption going forward
• Adoption trajectories differ by sector based on readiness of solutions, implementation
lag, and cost-benefit feasibility
• Productivity gains are expected to be redeployed at similar levels or with a reduction
of up to 20%
Exhibit 3
Results
The results indicate that accelerated adoption of AI across industries can contribute
$500B-$600B over and above India’s current GDP growth by 2035, driven by increased
productivity and efficiency in the workforce (Exhibit 4). The analysis shows that financial AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 19
services and manufacturing can be most impacted and might have up to 20-25% of their
sectoral GDP attributed to AI by 2035. Both sectors are detailed in the sections ahead.
Exhibit 4
Banking
AI-led productivity and efficiency improvement could unlock $50B-$55B in financial services,
over and above the current estimated growth for the sector by 2035 (Exhibit 4). This opportunity
will likely be realized as AI-mature Indian banks evolve into “bionic” organizations, combining
machines’ intelligence with humans’ judgment. AI accelerates business and functional
transformation across the banking value-chain, embedding intelligence into every product,
process and customer interaction. Reducing costs can also enhance financial inclusion.
Financial services companies’ front, middle and back offices are expected to be transformed
by machine learning and agentic AI. While the map represents important opportunities across
domains, areas with potentially the highest ROI have been highlighted (Exhibit 5):
• In the back office, AI could power automated compliance, fraud detection, and risk
management through advanced anomaly detection techniques and privacy-preserving
analytics such as secure multi-party computation and federated learning
• In the middle office, AI-enabled systems can reshape credit decisioning, collections,
and portfolio management. By leveraging alternative data sources, banks can make
more accurate, dynamic, and inclusive lending decisions
• In the front office, virtual relationship managers can deliver hyper-personalized
customer experiences. Using real-time behavioral predictions, these AI agents can offer
tailored financial advice, timely product recommendations, and proactive outreach,
helping deepen customer engagement and improve satisfaction across segments 20 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 5
...and a combination of AI & machine learning can transform financial services across
the front, middle, and back office
Front office Middle office Backoffice
Digital-led customer acquisition
Adaptive look-alike models scan daily click-
stream, score prospects, and auto-shift ad
budgets toward the highest-conversion
channels, boosting CAC efficiency.
Frontline sales enablement
Real-time call co-pilot transcribes the conversation,
matches needs to product bundles, inserts
mandatory compliance language, and logs next
steps straight into the CRM.
Relationship management and advisory
Generative assistant assembles concise
portfolio snapshots, flags risk or life-event
triggers, and drafts personalized action plans
for the relationship manager (RM) to approve
and send
Engagement, cross-selling, and customer
retention
Life-stage engine analyses transaction patterns and
sentiment to predict churn or upsell windows, then
launches hyper-targeted offers via push, email, and
RM dashboard.
Customer underwriting
Explainable ML combines bureau, cash-flow,
utility, and GST feeds to deliver real-time
affordability scores with clear reason codes
for the credit officer and the regulator.
Collections
The early delinquency model prioritizes overdue
accounts, picking the best channel, timing, and
repayment offer to maximize recovery at the lowest
collection cost.
Self-service through digital channels
Multilingual chatbot authenticates with
biometrics, handles KYC updates, disputes,
and card blocks end-to-end, and escalates
only edge cases to a human agent.
Assisted service
(contact center, branch, digital)
Voice analytics gauges sentiment and intent mid-
call, suggests relevant knowledge-base snippets,
while back-office bots auto-populate and route
service tickets.
Developer productivity
AI pair-programmer auto-writes routine
code, adds unit tests, tunes SQL queries,
checks for security issues, and flags build
problems before the code is merged. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 21
Potential enablers for consideration
Realizing the full potential of AI in India’s banking sector will depend on a set of enablers that
support innovation, adoption, and responsible scaling. These include:
Infrastructure
• Building capacity through Innovation Sandboxes to enable pilots focused on critical
themes such as explainable credit models, fraud and AML graph analytics, and self-
auditing regulatory technologies
• Utilizing a regulatory sandbox to test AI-related regulatory changes, e.g., video KYC
for NRI.
• Harness open, standardized dashboards from pilots to track business impact, fairness
outcomes, and emerging risks, enabling transparent supervision and learnings across
the ecosystem
• RBI has already announced the establishment of a public cloud infrastructure for the
financial sector
14
. Such efforts can be accelerated:
pAn empaneled set of vendors offering high-performance compute (e.g., GPUs),
privacy-preserving tools, and LLM-based APIs
pStrong data protection frameworks, including data replication, disaster recovery,
and compliance checks
pPre-approved toolsets for common banking use cases, made easily accessible
through a curated marketplace
• Define standards for and enable access to utility agents, trained on regulatory-grade
datasets and made available to banks and other financial institutions. RBI can define
standards to ensure these agents are explainable, compliant, and regularly updated to
reflect regulatory changes and market evolution
• Harness a cross-regulatory AI Innovation Sandbox to enable financial institutions to
test models in a secure environment alongside the regulatory sandbox.
• A shared “landing zone” for GPUs and computing resources on a pay-per-use basis for
smaller regulated entities, delivered via RBI or sector infrastructure providers.
• A dedicated funding corpus could support shared data and compute as public goods,
including grants for fintech accelerators and research labs.
• A sector-specific AI model for finance and make them available for safe adoption
across the sector.
14 Reserve Bank of India. Monetary Policy Statement. December 8, 2023. 22 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Data
• A central regulatory body for data governance in the financial sector. It could include,
but not be limited to, defining standards for classification of data, sharing across
entities, responsible use, security, and monetization.
• Defining frameworks (either enhancing existing or developing new ones) to include
alternative (especially unstructured) data sources, under a unified consent and
governance structure. An adequately authorized entity, such as the India Data
Management Office (IDMO), can standardize look at access issue to ensure borrower
opt-outs, and enable explainable scoring aligned with regulatory norms
• Harness existing data architecture frameworks to facilitate secure, anonymized data
exchanges between banks and other financial institutions for the purpose of model
training, ensuring privacy and transparency
• Broader adoption of alternative data by banks through enabling policies and common
infrastructure. An adequately authorized entity, such as the RBI, could:
pIssue clear guidelines on the use of such data for credit and other financial decisions
pEstablish a shared, consent-driven data repository, accessible to regulated entities
pPromote standardization and interoperability for secure data access and validation
• Consider whether a financial institution could maintain a regular inventory of its AI
systems, with anonymized metadata fed into a sector-wide repository for supervisory
visibility.
• AI can be more deeply integrated with existing digital public infrastructures (e.g., UPI,
AA, OCEN) to support enhanced public services such as multilingual access and fraud
detection.
Talent
• Training programs focused on the needs for the financial sector. As an example, a
national certification program in AI for Credit, Risk and Fraud could be considered
by the government, industry organizations, and universities. This would build trusted
and domain-aligned talent with expertise in responsible AI, explainability, and financial
regulation
• Consider whether tax incentives on course fees for financial institutions could increase
investment in skilling employees in AI, analytics, and cybersecurity. This would reduce
the post-tax cost for employers and spur large-scale adoption of upskilling programs
• A scalable and flexible talent pool through a national AI fellowship or exchange
platform:
pMaintain a vetted database of certified AI professionals available for temporary
deployments across banks, and other financial institutions, including regulators
pAllow small and mid-sized institutions to access top talent on-demand without full-
time hiring overheads AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 23
pFacilitate secondments or rotational programs across banks, RBI innovation arms,
and academia to deepen cross-sectoral expertise and rapid capacity building
Protection, Assurance and Governance
• Consumer protection and transparency by mandating fairness in AI outcomes, clear
disclosures on AI use, and accessible grievance redressal mechanisms, supported by
public reporting and toolkits that help smaller firms meet compliance.
• Resilience and security through continuous cybersecurity monitoring, dynamic threat
response, regular red-teaming of AI systems, and business continuity plans with
fallback mechanisms and drills.
• Governance and oversight by requiring AI-specific risk checks in product approval
processes, continuous monitoring of deployed models, institution-level AI inventories
feeding into a sector-wide repository, and risk-based audits (with independent third-
party audits for high-risk systems).
• Accountability through incident reporting and risk intelligence by establishing
tolerant, good-faith reporting mechanisms and aggregating disclosures to generate
sector-level insights on emerging risks.
Key risks
While AI in banking offers significant promise, India must also navigate a set of risks that
could impede adoption.
Data Infrastructure Talent Regulatory and IT Policy
Legacy IT core infrastructure
Most banks still operate on bulky legacy core
platforms. Integrating AI tools often requires
extensive rework or replacement, adding cost
and multi-year delays to realize productivity
benefits. These delays will persist without
sovereign AI/cloud infra and curated toolsets.
Data coverage and quality gap
Despite the push for Account Aggregator (AA) and
DEPA frameworks, many individuals remain outside
the digital data net, especially in low-digitization
areas. A lack of standardized and consent-driven
repositories further limits the availability and
reliability of alternative data for AI models.
Sandbox throughput constraints
Unlock the Innovation Sandbox’s capacity to
evaluate important interventions along the
financial services value chain. Without open
dashboards and dissemination of learnings,
systemic benefits will remain limited.
Privacy and consent fatigue
Consumer privacy concerns around alternative data
are intensifying. Without robust data protection,
explainability, and opt-out frameworks, public trust
may erode, constraining data access.
Lack of scalable AI utility infrastructure
Absence of centralized AI tools, agents,
and vetted vendors for common banking
use cases forces every bank to reinvent the
wheel. This fragmentation hinders innovation
velocity and increases cost.
Unequal access to AI talent and tools
Smaller banks and fintechs may lack the resources
to tap into centralized utility agents, high
performance compute environments or rotational
AI pools, reinforcing a digital divide within the
financial sector.
Upskilling & change management
Without clear roles, staff training, and
guardrails, using AI agents can lead to errors,
compliance issues, and serious risks to
customer trust and system stability. 24 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Manufacturing
In manufacturing, $85–100B could be driven by AI-led productivity and efficiency improvement
over and above India’s current growth by 2035. The National Manufacturing Mission outlines
five key pillars
15
: Ease of Doing Business, Future-ready Workforce, Vibrant MSME Sector,
Availability of Technology, and Quality Products, of which AI will have a high impact on three:
Availability of Technology, Future-ready Workforce, and Vibrant MSME Sector.
AI can unlock productivity and efficiency across multiple dimensions by lowering the cost
of production, improving output yields through enhanced process efficiency, and increasing
throughput via predictive maintenance on the shop floor. It can also enable the production
of higher-quality goods at similar prices by powering intelligent product design, real-time
quality control, and mass customization. To fully realize these benefits and build a future-
ready, competitive industrial base, upskilling India’s manufacturing workforce in AI tools will
be essential (Exhibit 6).
For India to fully capture the gains from AI-native manufacturing, it is important to strengthen
both forward and backward linkages. On the backward side, this means building resilient
supply chains, integrating AI-ready MSMEs, and ensuring reliable access to inputs. On the
forward side, India can actively expand domestic markets and position itself in global value
chains through coordinated industrial and trade policies. Productivity gains alone will not
deliver impact unless industrial policy, trade strategy, and demand generation evolve together
to convert efficiency into competitiveness and growth.
Exhibit 6
15 Press Information Bureau, Government of India. “‘National Manufacturing Mission’ to cover small, medium and large
industries for furthering ‘Make in India’ announced in Union Budget 2025-26.” February 1, 2025 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 25
Potential enablers for consideration
Realizing AI’s full potential in India’s manufacturing sector will depend on enablers that
enhance productivity and foster innovation-driven growth at scale.
Infrastructure
• “AI-Ready Industrial Parks” that place clean-energy factories next to high-performance
computing (HPC) labs and skills centers, similar to the technology-and-fabrication
clusters promoted under the United States CHIPS and Science Act
• Government decision makers to look at -funding shared facilities for 3-D printing,
advanced materials testing and precision metrology can give MSMEs affordable access
to expensive equipment
• Robotics and humanoid test zones with safety rules that allow rapid prototyping,
already being done in several Chinese electric-vehicle (EV) hubs, could speed up local
innovation in collaborative robots and vision-guided welding stations
• A national BharatNet fibre backbone with Time Sensitive Networking (TSN) - capable
switches will let control signals travel reliably between separate buildings or even
distant factories that share the same production line
• An online SaaS marketplace where Indian and global developers can publish plug-and-
play AI tools, for example, for anomaly detection or energy optimization, which will let
factories subscribe to the exact algorithms they need instead of investing heavy capital
upfront
Data
• An open “Manufacturing Data Grid,” a shared platform where OEMs, suppliers, other
stakeholders, and startups can trade production and supply-chain data through
standard APIs, taking inspiration from Germany’s Manufacturing-X data-space model
• Industrial parks to house both cloud servers and edge computers equipped with GPUs
so that high-speed quality checks, digital twins and predictive-maintenance systems
can run close to the machines that generate the data
• Industrial clusters with 5G and 6G mobile networks combined with Time-Sensitive
Networking (TSN) switches so robots and Automated Guided Vehicles (AGVs) can
send and receive data without delay
• Data sharing framework for manufacturing so that companies can share sensitive
design drawings or process settings only with trusted partners and always with clear,
revokable permissions
• National libraries (or marketplaces) of reusable digital-twin models, for example, for
semiconductor fabrication plants, battery-cell lines, precision auto-parts machining
and aerospace composites to shorten design cycles without starting every simulation
from scratch
Talent
• The Ministry of Skill Development, the All-India Council for Technical Education
(AICTE) and leading manufacturers could explore how to launch a tiered “AI for
Advanced Manufacturing” program that starts with micro-badges and scales up to full 26 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
postgraduate programs, following Japan’s Ministry of Economy, Trade and Industry
(METI) model for rapid AI upskilling
16
• Industry-academia research chairs in specific domains, e.g., chip design, power
electronics, robotics and battery chemistry to allow professors and graduate students
to work on real-world factory challenges while keeping intellectual property inside
India
• Government decision makers could assess the benefits of a targeted reverse-diaspora
program that offers fast-track visas, research grants and senior leadership roles to
encourage experienced semiconductor, battery and automation experts living abroad
to return and teach or build in India
• A national registry of certified freelance AI-maintenance engineers, digital-twin
modelers and automation troubleshooters that will help factories scale their specialist
workforce up or down
Key risks
Risks that could hinder India’s push toward AI-native manufacturing include
Data Infrastructure Talent Regulatory and IT Policy
Fragmented data ecosystem
Without a standardized Manufacturing Data
Grid and common APIs, production and
supply-chain data will remain siloed across
OEMs, MSMEs, and startups. This limits AI’s
ability to deliver end-to-end visibility and
optimization.
Inadequate edge + network infrastructure
AI use cases like digital twins or predictive
maintenance need low-latency computing close
to machines. Many parks lack edge GPUs, 5G/6G,
or TSN-capable networks slowing adoption.
Shortage of cross-skilled talentShortage of cross-skilled talent
India’s AI workforce has room to deepen
expertise in manufacturing (e.g., robotics,
chip design, battery tech). Upskilling via tiered
credentials and reverse diaspora programs
is critical but currently limited in scale and
alignment with real factory needs.
Low awareness and access to shared AI
infrastructure
MSMEs often can’t access HPC labs, 3D printing
centers, or plug-and-play SaaS AI tools due to
cost, geography, or lack of knowledge, widening
the gap between large OEMs and small suppliers.
Slow adoption of DEPA-like consent
protocols
Sharing sensitive factory data (e.g., designs,
process settings) is essential for collaborative
AI. Without robust consent-based, revokable
frameworks, companies may avoid data
sharing due to IP fears.
Slow industry readiness tracking
Without government support and structured
assessments like SIRI, many MSMEs and smaller
organizations may struggle to identify critical
skill or technology gaps—slowing their ability to
scale and transition toward Industry 5.0.
AI adoption across industries represents a critical lever for India to enhance productivity
and competitiveness, with banking and manufacturing standing out as early opportunities.
Building on these foundations, the next chapter explores Lever 2: Unlocking leapfrog
innovation by transforming R&D with AI—a pathway for India to accelerate discovery, shorten
innovation cycles, and establish a stronger foothold in global, innovation-driven industries.
16 Ministry of Economy, Trade and Industry (METI), Approaches to human resources and skills required for DX
promotion in the age of generative AI, August 7, 2023. AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 27
CHAPTER 4 – LEVER 2: UNLOCK LEAPFROG INNOVATION BY
TRANSFORMING R&D WITH AI
Transforming R&D, especially using generative AI, can enable India to leapfrog into innovation-
based opportunities on a global scale by transcending traditional growth pathways and
creating new products, services, and business models. Historically, such innovation paths have
been challenging for India due to the heavy capital required for conventional R&D, e.g., $1–2B
17
for novel drug development and $1–1.5B to engineer an all-new electric vehicle platform. AI
lowers these entry barriers. It can accelerate commercialization, disrupt legacy value chains,
and create a lasting competitive edge. Analysis suggests that such breakthrough innovations
could potentially contribute at least an incremental $280-475B to India’s GDP by 2035.
Approach
11 sectors have been selected from McKinsey’s 18 global arenas
18
of growth and 7 additional
India-specific arenas based on their local relevance and transformative potential.
• Baseline (2023) revenues across all 18 arenas were sourced from current industry reports
and total $640–750B
19
. Projections for 2030 were developed using a combination of
industry estimates and expert consultations (e.g., ONDC for next-gen e-commerce),
reaching $1.7–2.0T
• The 2035 revenue projections were developed using 2023–2030 growth trends,
resulting in an estimated $3.6–4.5T range. To assess the incremental GDP contribution,
the revenue delta from 2023 to 2035 was calculated and sector-specific revenue-to-
GDP conversion factors were applied, resulting in a projected GDP impact of $1.4–1.9T
across all 18 arenas by 2035
• Out of 18 arenas, 10 were identified where AI is the primary value driver for leapfrog
growth. These 10 arenas are expected to contribute $380–660B in incremental GDP
impact by 2035
• This yields an estimated $280-475B in incremental GDP attributable specifically to AI-
led leapfrog growth across the 10 arenas by 2035
The exhibit below shows the identified 18 arenas of growth that could serve as key drivers of
India’s growth over the next decade, with a projected GDP impact of $1.4–1.9T by 2035.
17 World Economic Forum article titled “How blockchain can cut the cost of new medicine”. December 2018.
18 McKinsey Global Institute report titled “The Next Big Arenas of Competition”. October 23 2024 – 18 global arenas of
which 11 are directly relevant for India and 7 additional were added
19 McKinsey article titled “India’s Future Arenas: Engines of Growth and Dynamism”. June 19 2025. 28 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 7
Of the 18 arenas of growth, ten are primarily driven by AI-led innovations such as AI software
and services, a global hub for auto-components, cloud services, software-assisted vehicles,
semiconductors and PCBs, medical devices, SpaceTech, aerospace and defense, cybersecurity,
and biopharma (see exhibit below).
Exhibit 8
It is important to recognize that AI-led growth will be shaped by unexpected breakthroughs
that remain beyond our foresight today. AI is advancing at an extraordinary pace with AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 29
AI supercomputing capacity doubles every 8–10 months.
20
, and AI algorithmic efficiency
doubles every 15–18 months
21
. Given this velocity, it is difficult to anticipate the full spectrum
of innovations that may arise. While 18 opportunity industries for India have been identified,
the rapid evolution of AI could well unveil a 19th or 20th frontier that will likely emerge over
time.
The next section illustrates the AI-led opportunity in the pharmaceuticals and automotive
sectors, which are aligned with India’s factor endowments.
Pharmaceuticals
Currently, 80% of the Indian pharmaceutical market is driven by generics
22
. This is because
the high costs of developing a novel drug (up to $1–2B per molecule)
23
, long timelines (over
10 years)
24
, and significant financial risks have historically limited investment in innovative
R&D capabilities. Emerging technologies such as AI can help lower development costs and
timelines across the drug discovery and development value chain, enabling India to transition
from generics to the innovator space over the next decade. India’s expertise in generics,
domain talent (e.g., pharmacology) and its endowment in the form of a rich genetic pool can
position it well to capture this opportunity.
Traditional drug development is divided into five distinct stages and typically takes >10 years
to complete end-to-end, with a potential capital spend of $1–2 billion (Exhibit 9).
Exhibit 9
20 arXiv preprint titled “Trends in AI Supercomputers”. April 22 2025
21 Forethought Research article titled “Will AI R&D Automation Cause a Software Intelligence Explosion?” March 25
2025
22 IQVIA market data
23 World Economic Forum article titled “How blockchain can cut the cost of new medicine”. December 2018
24 N-SIDE blog post titled “What’s the average time to bring a drug to market in 2022?”. November 5 2022 30 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
AI is re-shaping the traditional drug discovery value chain. It could reduce R&D costs by
20–30%
25
through drug repurposing, AI-driven research and documentation, and replacing
traditional placebo groups in clinical trials with AI-generated virtual placebos. This can
simulate control groups without needing real participants; shorten drug discovery timelines
by 60–80%
26
via AI-powered molecule design and Insilico modelling to speed up lead
identification by four times; improve clinical trial success rates by 5–15% by leveraging India’s
diverse gene pool to identify optimal patient subgroups.
Additionally, frontier technologies are creating new pathways for drug development. Platform-
based approach is enabling the creation of reusable, end-to-end technology engines that
can discover, design, and optimize multiple drugs across therapeutic areas by integrating
AI models, large-scale genomic data, and automated lab systems. One key advantage of
this approach is the ability to repair or retune existing molecules in days or weeks when a
pathogen or tumor mutates, avoiding the need to restart the discovery process from scratch
For example, a Boston based startup advanced a novel lung fibrosis drug from concept
to Phase 1 trials in under 30 months, while another Canada-based startup developed a
COVID-19 antibody in just 90 days using its AI-driven antibody discovery platform.
India could consider licensing and launching 90–110 innovative drugs by 2034 across four
phases: post-molecule discovery, post-phase 1, post-phase 2, and E2E commercialization.
This would culminate in a value capture of $5–8B and establish India as an innovation-led
hub (Exhibit 10).
Exhibit 10
25 McKinsey report titled “Generative AI in the pharmaceutical industry: Moving from hype to reality”. January 9 2024
26 McKinsey report titled “Generative AI in the pharmaceutical industry: Moving from hype to reality”. January 9 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 31
Potential enablers for consideration
Realizing value in pharmaceuticals could depend on enablers to improve clinical research,
optimize manufacturing and ensure regulatory readiness.
Infrastructure
• Expanded the number of biotech parks by 10x to support the predicted expansion in
research and development led by the growth of startups
• High-performance computing infrastructure in line with EU and China to meet the
intensive computational needs of AI-driven drug discovery
Data
• Diverse genomic and clinical datasets to build a unified, high-quality national omics
dataset, led by institutions such as the Indian Biological Data Centre (IBDC) and
Biotechnology Industry Research Assistance Council (BIRAC)
• A tiered data access model like the U.S. National Institutes of Health’s “All of Us”
program, providing public, registered, and controlled access levels to researchers
27
Talent
• Train and retain biopharma R&D scientists by 2035 to support the annual development
of 20–25 new drugs (Exhibit 10)
• Undergraduate and postgraduate programs in computational biology and bioinformatics
in collaboration with premier technical institutes
• Expanded central research grant pools to have a dedicated track for AI in drug discovery
that focuses on attracting global talent
• Enhanced re-entry fellowships with increased stipends and research grants to match
global talent programs
Policy and Regulations
• Government decision makers could review India’s pharmaceutical regulations to ensure
they align with global standards to facilitate faster clinical trials and international
recognition
• Government decision makers could explore the potential benefits of a data exclusivity
law that could protect clinical trial data while incentivizing innovation
• Government decision makers could explore whether there is scope to streamline the
clinical trial approval process, a potential 30-day approval route for institution-initiated
trials would match global best-in-class timeframes
• Government decision makers could look at how to implement global best practices
for vaccine approvals, such as rolling data reviews and digital submissions that aim to
shorten overall timelines for time to market
27 All of Us Research Hub article titled “Data Access Tiers” 32 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Market Access
• A drug-access acceleration fund through BIRAC to co-invest in the commercialization
of Indian therapies
• A “Made-in-India Innovation Track for Pharma” to mandate the listing of breakthrough
drugs within 180 days of approval
• Empower pre-revenue biotech firms to go public by mirroring NASDAQ’s non-revenue-
based listing parameters
• Consider the potential benefits of offering capital incentives, such as capital grants for
AI-R&D centers and reduced tax rates on profits from India-patented drugs
• Global pharmaceutical firms with priority access to national high-performance
computing infrastructure and a 180-day fast-track approval path for AI-designed drugs
Key risks
India’s goal of AI-based drug development faces potential risks
Data Infrastructure Talent Regulatory and IT Policy Market access
Value erosion
U. S . market reforms could cut global drug out-
licensing value
28
, sharply curbing the earnings
potential of Indian-origin therapies.
Slow and uncertain domestic market
Unless the pharma regulatory body shortens
drug-approval timelines from 20–25 months
to 15 months, matching global benchmarks,
each year of delay can erode 10-15% of an
asset value.
29
.
Unclear guidelines on AI-discovered drugs
Global patent norms requiring “significant
human contribution” may limit protection for
AI-generated molecules, risking rejection and
long-term barriers to global commercialization
in key markets.
Genomic data gap
Reaching the omics dataset target by 2035
faces major risks, high capex, potential legal
backlash over consent, and low participation
due to rising privacy concerns.
High-performance compute bottleneck.
AI in drug discovery requires dedicated HPC
capacity, but risks like lack of funding, global
GPU shortages, 70+ week lead times
30
, and
export controls could delay deployment by
18–24 months.
R&D talent gap
India’s low researcher density and
uncompetitive stipends may hinder progress
toward the 2035 target of 1.5k researchers per
million, unless talent training and retention
improve.
Automotive
AI can emerge as a game changer for the automotive sector in India, enabling it to cut
costs, improve safety and accelerate innovation. In the following sections, the report
explores two pathways for automotive: Software-Assisted Vehicles (SAVs) and AI-enabled
component design. Harnessing frontier technologies, including RFID-based smart corridors,
5G-connected routes, and AI-driven design and validation, could put 18-20M software-
ready vehicles on Indian roads by 2035 and unlock $20-25B in export gains and import
substitution. (Exhibit 12 and Exhibit 14)
28 OHE bulletin titled “US drug pricing policies will have global impacts on innovation and access”. June 30 2025
29 McKinsey article titled “The road to positive R&D returns”
30 Industry Technology Report titled “Prepare for the Coming AI Chip Shortage”. September 25, 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 33
Software-assisted vehicles (SAVs)
Software-Assisted Vehicles (SAVs) represent the next generation of automobiles, where
core functionalities are increasingly driven by software rather than hardware-intensive
systems. SAVs operate across five defined levels of autonomy, as per the Society of
Automotive Engineers (SAE) International (Exhibit 11). The automotive industry in India
is presently concentrating its efforts on progressing from Level 2, which features partial
vehicle autonomy, towards achieving Level 4, characterized by highly autonomous driving
capabilities. These vehicles rely on flexible electronic architectures, connected systems, and
over-the-air (OTA) updates to minimize human intervention.
India is expected to reach Level 3 by 2035, with its AI-led automotive inflection point
between Levels 3 and 4. As India emerges as a major SAV consumer market and global
production hub, this shift offers a key opportunity for domestic value creation and global
competitiveness.
Exhibit 11
By 2035, 40-50% of the total 40-45M vehicle market, i.e. 18-20M units, would be enabled
via software (Exhibit 12). These will be split across passenger vehicles at 4-5M, commercial
vehicles at 1-2 M and two-wheelers at 13-15M. India could unlock $6-8B cumulative domestic
value through AI-enabled SAV domestic subscriptions by 2035, with an estimated $1.5-2B
annual exit value that year. 34 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Exhibit 12
A three-level autonomy-driven pathway for India’s SAV evolution is considered. India’s SAV
evolution will move from Level 2 partial automation (driver-assisted) to Level 3 conditional
automation (hands-off in specific scenarios) and Level 4 high automation (self-driving in
geofenced areas).
Unlocking opportunities beyond conventional technologies
Conventional autonomous driving relies on costly on-board sensors like LiDAR, cameras,
and radars combined with real-time AI processing to navigate roads without human input.
However, such systems may face challenges in India due to weather or traffic conditions
and poor road markings. India can explore alternate, infrastructure-assisted approaches to
enable affordable and reliable autonomy, including:
• RFID-based corridors that help vehicles localize accurately in all weather, reducing
reliance on GPS or cameras
• Magnet marker guides lanes, which act like a virtual rail, enabling lane-keeping, even
on waterlogged roads
• 5 G-enabled corridors that share real-time data over 5G, helping vehicles detect hidden
traffic in urban settings
• Satellite navigation system in combination with real-time kinematic from ground towers
provides centimeter-level accuracy ideal for remote areas (Exhibit 13). AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 35
Exhibit 13
To unlock $6-8B in SAV value over the next decade, AI can help accelerate engineering
capability building and faster time to market for software-enabled automotive products at
reduced costs. AI copilots or custom LLMs trained on AUTOSAR/ vehicle OS documents
can cut learning time by 30–50%. AI-powered model-based design can reduce design cycle
times by 20–30%, while AI-led bug detection and regression testing can lower validation
costs by up to 40%. Additionally, AI vision and reinforcement learning can optimize compo-
nent assembly and reduce Electronic Control Unit (ECU) testing costs.
Potential enablers for consideration
Realizing the full potential of shared autonomous vehicles in India will depend on enablers
that build robust digital infrastructure, ensure safety and regulatory readiness.
Infrastructure
• 6–8 physical-digital testing parks
31
by 2035 to validate autonomous driving features in
Indian conditions
• 10,000 km
32
of 5G/early-6G corridors with roadside units to enable real-time data
exchange for software-assisted vehicles
• Designated 10-20 low-risk deployment zones by 2030 in areas such as airports and
campuses for the initial commercial rollout of Level 3 and Level 4 autonomous vehicles
31 Benchmarking with testing parks in South Korea and U.S. per million cars; American Center for Mobility report titled
“About the American Center for Mobility”; K-City report titled “K-City: South Korea’s 5G-Connected Autonomous
Vehicle Testbed”
32 Covering Golden Quadrilateral, connecting Delhi, Mumbai, Chennai, and Kolkata (~5,900 km); high-impact national
corridors (~2,000 km), including Ahmedabad–Nagpur, Mumbai–Bangalore, and Delhi–Kolkata; and top 20 urban
ring roads (~2,000 km) 36 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• Three regional Zero Prototype Labs in key automotive component hubs e.g., Pune,
Aurangabad that provide digital testing and simulation facilities, particularly for MSMEs
and startups
Data
• Incentivized manufacturers sharing anonymized telemetry data from 20-25% of
new vehicle sales annually to create a large-scale national sensor dataset. Utilize the
collected data to define safety norms for semi-autonomous vehicles and identify high-
risk roads for infrastructure improvements
• A centrally owned and shared technology stack for Software-Assisted Vehicles (SAVs),
including standardized hardware, software interfaces, and safety regulations to reduce
costs and ensure interoperability, an open SAV reference architecture and a national
high-definition map cloud through public-private partnerships
Talent and Capability Development
• Train and retain a workforce of 30,000+ engineers
33
that specialize in SAVs by 2035
• Centers of Excellence in vehicle software at premier engineering and management
institutes, and introduce SAV-focused minor degree programs
• A Global Mobility Tech Visa to attract international talent with streamlined processing
and competitive compensation
• Upskill and retain engineers with combined expertise in AI and mechanical engineering
by 2035 to support the auto components industry
• Open public-private Centers of Excellence to reskill OEM engineers, integrate AI into
undergraduate curricula, and host industry-supported hackathons
Supplier Ecosystem
• Grow the SAV-ready supplier base to over 100 firms
34
across sensors, ECUs, chip
design, and AI services
• Two auto-grade semiconductor fabs with a monthly production capacity of 40,000+
wafers
35
under the India Semiconductor Mission to meet domestic demand and
potentially create an export surplus
33 Benchmarking against Volkswagen’s 10k engineers for 9M cars and scaling down to 0.6M vehicles/year for an Indian
OEM (i.e., 660 engineers), taking a 20% cut for in-house engineers, yields 100–120 engineers/ OEM for India’s 38-40
OEMs
34 Benchmarking with Germany’s SAV-ready suppliers, scaled to India’s aspiration of producing 5M+ vehicles/year
by 2035; GTAI (Germany Trade & Invest) report titled “Automotive Industry – Germany’s Production of 4.1 Million
Passenger Cars”. October 10 2024.; Meyer Industry Research report titled “TOP 100 Automotive Suppliers Germany
2021”
35 EE Times article titled “ESMC 300-mm Wafer Fab: A Bid to EU’s Semiconductor Sovereignty”. November 7 2024 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 37
Regulatory
• Government decision makers could review standard frameworks for vehicle
cybersecurity and over-the-air software updates to align with global regulations
• Government decision makers could seek to form an SAV Regulatory Taskforce under
the Ministry of Road Transport and Highways to develop agile and globally aligned
standards
• Government decision makers could look at the potential of a fast-track AI patent
regime to reduce the patent grant timeline to under 20 months
Market Access and Compliance
• Concessional GST slab or an income-tax deduction on financing costs for components
designed or validated using AI e.g., Deep-Learning Surrogates (DLS) to encourage
adoption
• “Digital Patent Box” with a lower tax rate for licensed DLS models and AI workflows
developed and commercialized in India
• Global AI safety standards integrated into the existing AIS-140 framework to ensure
compliance and certify AI-generated components for domestic and international
markets
• National Surrogate-Model Certification Sandbox established, e.g., at ARAI Pune, to test
and certify AI models for the automotive industry
Key risks
India’s aspiration of software vehicles is prone to following risks
Data Market access Talent Supplier ecosystem Regulatory and IT Policy
Regulatory misalignment
If the automotive sector does not harmonize
its cybersecurity and software-update rules
with UNECE WP.29 R155/R156 by 2028,
Indian SAVs could fail EU/Japan official
certifications, shrinking exports
36
Domestic supplier base stagnation
If component suppliers do not innovate
quickly enough to develop modular, certified
parts, the penetration of SAVs will remain
limited - slowing roll-outs and constraining
revenue growth
Semiconductor shortfall
If two auto-grade 28/16nm fabs are not
commissioned & global chip procurement
lead time continue to remain high (currently
>70 weeks), assembly lines can idle for 3-5
quarters, reducing value capture
Engineering talent gap
If only 25-30k of the required 35k-40k SAV
engineers are trained or attracted, software
cycles may lengthen 6-12 months, trimming
value capture from subscription revenues
Cyber-trust shock
If a major cyber hack occurs before robust
cybersecurity compliance is in place, it can
crash consumer trust and limit adoption rates.
Feature non-democratization risks
Unless industry bodies actively democratize
and scale Tier 1 and Tier 2 SAV features,
adoption could stay far below the 40–50%
target for 2035
36 UNECE press release titled “Three landmark UN vehicle regulations enter into force”. February 5, 2021 38 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Auto components design
AI-powered models such as Deep-Learning Surrogates (DLS) replicate the behavior of
complex physics simulations, enabling near-instantaneous and highly accurate predictions.
These models are transforming the R&D value chain, redefining how components are designed,
tested, and manufactured. Traditionally, simulating component behavior like aerodynamic
drag, thermal stresses, or structural deformation requires extensive physics-based computing
that can take hours to days per iteration. Once trained, the AI models perform the simulations
in milliseconds, significantly accelerating R&D cycles and enabling more efficient, low-cost
innovation.
For India, which currently accounts for 1-2%
37
of the $500 to $550B
29
global automotive
parts export market i.e., $7-8B, DLS can be an enabler to increase its share significantly. With
automotive parts imports at $6-7B
29
in high-potential areas, there is an opportunity to reduce
import dependency by improving domestic design and testing capabilities. AI-led design,
including DLS, not only boosts competitiveness by slashing development time and costs but
also allows India to lead in high-value, design-driven exports moving beyond assembly and
manufacturing. Indian auto components OEMs can potentially capture $25-30B in cumulative
value by 2035 with an exit value of $4-6B in 2035 (Exhibit 14).
Exhibit 14
37 United Nations Comtrade Database titled “UN Comtrade: International Trade Statistics Database” AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 39
Potential enablers for consideration
Enablers that enhance skill development, strengthen compliance and improve overall
performance could benefit the components sector.
Infrastructure
• Automotive bodies can jointly establish three regional Zero Prototype Labs (ZPLs) in
India’s key component hubs - Chennai/Tamil Nadu (35% revenue share), Pune/Western
belt (33%), and Gurgaon-Manesar (30%). These ZPLs would serve as digital testbeds using
AI, digital twins, and virtual simulations to eliminate early physical prototyping. They can
reduce development time by 20–25% and costs by 10–30%. While large manufacturers
can set up in-house ZPLs, MSMEs and startups will benefit from shared facilities
• Examples of similar lab setups globally include EDAG’s Zero Prototype Lab in Wolfsburg
38
(~1,700 m, 7-9 petaflops of compute power, run 4.5-6K parallel simulations) and Porsche’s
center in Weissach
39
(~2,100 m, 10-12 petaflops of compute power, run 6-8K parallel
simulations)
• The labs can be potentially run on a shared access model with access defined across
three tiers:
pCorporate tier: Paid, high-priority access with dedicated compute and IP security
pMSME & startup tier: Subsidized, affordable simulation slots
pCoE tier: Open academic access for public R&D and skill-building
Incentives and IP Frameworks
• IP for Deep-Learning Surrogate (DLS) models spans data, model architecture, and
digital twin workflows. With significant model training costs, legal protection is key to
driving investment. India could strengthen its incentives and IP protection frameworks
by creating a fast-track AI patent regime to reduce the current patent grant timeline
from 48+ months
40
to less than 18 months, ensuring innovators can secure IP rights
swiftly before data or model designs are leaked or copied
• The shorter timeline would close India’s “protection gap” with the U.S. (about 23 months,
with a 12-month “prioritized” track for digital technologies) and South Korea (standard
16 months)
Compliance
• Physics-Informed & Explainable AI (PI-XAI) means AI models used in vehicles should not
only be as accurate as traditional simulations like Computational Fluid Dynamics (CFD)
but also follow basic physical laws and clearly explain their predictions. For instance,
a 2024 study showed a physics-based AI model could predict battery health with just
0.87% error
41
. But experts highlight that regulators still need clear safety limits and ways
to measure uncertainty before certifying these models.
38 EDAG article titled “Zero Prototype Lab” (part of “Reducing time-to-market”). November 21 2023
39 Porsche Newsroom article titled “Driving simulators: Test drive without a test vehicle”. May 2 2025
40 Times of India report titled “At over 90k patent filings, highest in 2 decades”. January 2024
41 Nature Communications article titled “Physics-informed neural network for lithium-ion battery degradation stable
modeling and prognosis,” Published May 21, 2024 40 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
• To encourage compliance, India could integrate global AI safety standards like ISO/PAS
8800:2024 (for explainability and transparency in vehicle AI) and ISO/ IEC 42001 (for
ethical AI management) into its existing AIS-140 framework
42
. This would help suppliers
certify AI-generated components for both Indian and global markets. This is in line
with Germany’s Technischer Über-wachungsverein Süd (TUV SUD) (German technical
inspection and certification organization), which already runs ISO 8800-aligned “AI
Quality” audits for automotive suppliers
43
• Setting up a National Surrogate- Model Certification Sandbox (NSCS) at ARAI Pune
to test and certify AI models using open datasets, physics checks, and cybersecurity
setups, feeding results into the new AIS-XAI approval process.
• The above requirements are not exhaustive; industry organizations could lead the
collation of a complete set of relevant global standards and certifications, and ~3
months under its accelerated route for semiconductor and Artificial Intelligence filings
Talent
• Large Indian auto-component manufacturers can build dual learning tracks, micro-
credentials in DLS for all R&D engineers, and bootcamps for business-unit heads.
This mirrors a global auto components manufacturer’s upskilling of 130K+ employees,
and AI leadership programs at leading automobile OEMs, which have reached 50K+
managers. Indian automotive organizations can establish 20+ public-private Centers of
Excellence (CoEs) hosted in IITs, NITs, IISc, and IIMs/ISB to:
pReskill engineers via semester-long residencies
pEmbed AI–mechanical modules in undergraduate curricula
pRun Formula-Student style hackathons using DLS-only validation, supported by
industry datasets and grants
• Draw inspiration from global automobile OEMs that have collaborated with technical
universities to foster joint research in AI and autonomous systems, or have focused
on research in sensor and AI integration, which helps automotive firms build in-house
expertise in surrogate modelling and AI-driven simulation
42 ISO/PAS 8800:2024 functional-safety standard for AI in road vehicles. Published December 2024; ISO/IEC
42001:2023 AI-Management-System standard. Published December 2023
43 TÜV SÜD AI Quality Certification Program (AIQCP) training programme. (2024) AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 41
Key risks
India’s aspiration to be AI first automotive component manufacturer is prone to risks.
Data Infrastructure Talent Market access Regulatory and IT Policy
Certification risk
If India auto manufacturers doesn’t launch a
certification sandbox and align with global
AI safety norms, auto parts may miss key EU/
US approval windows, causing 1–2 year delays
and risking export contracts.
Infrastructure gap
If the three Zero Prototype Labs aren’t set up by
2029 due to funding gaps, validation will shift
back to physical testing, slowing time-to-market
by 20–30%, potentially eroding value gains.
Supply chain adoption risk
If MSME suppliers can’t access “DLS-as-a-
service” tools, due to high licensing fees, or
limited technical support, overall adoption
may remain low, potentially limiting India’s
potential value capture.
Geopolitical risk
If future WTO or India–EU rules require firms
to disclose training data for safety audits, &
companies see this as a threat to their IP, they
may limit model exports, shrinking India’s global
market access.
AI model integrity risk
If physics-informed quality checks and secure
update systems aren’t in place, even one
major prediction error or cyberattack could
lead to global recalls and a freeze on using
AI-designed parts. 42 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
WAY FORWARD FOR THE INDUSTRY
The full potential footprint of AI on India’s economy is difficult to visualize, and this report
offers a starting point. It outlines two levers with two priority areas each, but the same
structured approach could be extended to other sectors such as logistics, construction, and
retail. Global benchmarks already illustrate the scale of opportunity: the IMF estimates AI
could lift global GDP growth significantly over the next decade
44
; the World Economic Forum
projects that nearly a quarter of all roles worldwide will change within five years due to AI
adoption
45
; and McKinsey research suggests that generative AI alone could contribute more
than any single technological wave in recent history
46
. These underline the importance of
applying an India-specific perspective to additional sectors to identify use-cases, value pools,
and enabling conditions.
Equally, labour transitions would be central to how India adopts AI. International institutions
estimate that around 35-40% of jobs worldwide are exposed to AI, with higher exposure in
advanced economies and meaningful effects across emerging markets
47
. Projections show
that while AI will create many new roles, it will also displace many existing jobs, particularly
in clerical, routine, and low-skill segments. For India, the challenge would be twofold:
preparing a workforce with advanced digital and AI skills to capture new opportunities,
while simultaneously ensuring that those displaced are gainfully employed through reskilling,
redeployment, or absorption into other growth sectors of the economy.
Finally, productivity gains and innovation must match market creation to translate into
growth. India would need to simultaneously deepen domestic demand and secure stronger
participation in global value chains. This will require alignment of industrial and trade policies,
particularly as global rulebooks evolve quickly. For instance, the European Union’s AI Act will
phase in obligations for general-purpose and high-risk AI systems, and new climate-related
trade measures such as carbon border adjustments are set to shape market access conditions
48
.
How India anticipates and responds to global shifts will influence its competitiveness, its
ability to attract investment, and its standing as a credible global partner in the AI economy.
44 IMF working paper titled “The Global Impact of AI: Mind the Gap”. April 2025
45 World Economic Forum press release titled “Future of Jobs Report 2023: Up to a Quarter of Jobs Expected to
Change in Next Five Years”. April 2023
46 McKinsey report titled “The Economic Potential of Generative AI: The Next Productivity Frontier”. June 14, 2023
47 IMF blog post titled “AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity.” January 14, 2024
48 Website titled “EU Artificial Intelligence Act | Up-to-date developments and analyses of the EU AI Act.” 2025 AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 43
APPENDIX
Seven Pillars of the IndiaAI Mission
49
Compute
capacity
Establish a federated public infrastructure of high-end GPU clusters
and launch AIRAWAT (AI Research Analysis and Workbench). As of 15th
September 2025, the mission targets deployment of 38k+ GPUs to support
national compute requirements
Innovation
centre
Develop India-specific foundational models and large language models
(LLMs) tailored to Indian languages and domains, advancing IP creation
Datasets
platform
Build anonymized, consent-based, interoperable public datasets to
fuel model training. This platform underpins innovation, application
development, and trusted AI governance
Application
development
initiative
Build AI solutions for agriculture, healthcare, education, and mobility
through targeted pilots and partnerships with industry.
Future skills
Train 50L+ students and professionals through AI curriculum integration,
data labs, and new certification programs in higher education
Startup
financing
Support 1,000+ AI startups through catalytic funding for deep-tech and AI
product innovation
Safe and
trusted AI
Create frameworks and toolkits to ensure explainable, ethical, and privacy-
preserving AI. This includes audit tools and compliance standards.
49 IndiaAI portal (Government of India initiative) 44 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Glossary
AI: Artificial Intelligence: In this report, AI
includes software capabilities such as machine
learning, data analytics and hardware-driven
automation, such as robotics
GDP: Gross Domestic Product: The total value
of all goods and services produced in a country
in a given time
BAU: Business as Usual: A scenario where
current trends continue without major policy
or technology changes
MSME: Micro, Small, and Medium Enterprises:
Businesses that are classified based on
investment size and number of employees
E2E: End-to-End: A complete process or
system that functions seamlessly from start to
finish
AA: Account Aggregator: A system that allows
secure and consent-based sharing of financial
data between institutions
IDMO:
India Data Management Office: A
body responsible for overseeing non-
personal and anonymized data governance
in India
DEPA: Data Empowerment and Protection
Architecture: A framework enabling users
to share personal data securely with
consent
RBI: Reserve Bank of India: India’s central
banking institution that regulates monetary
policy and ensures financial stability
RBIH: RBI Innovation Hub: A unit under RBI
supporting fintech and technology-led
innovations in the financial sector
ReBIT: Reserve Bank Information
Technology Private Limited: An RBI-owned
tech company managing cybersecurity
and IT infrastructure
AML: Anti-Money Laundering: Regulations
and systems designed to prevent illegal
money transactions and funding sources
GPU: Graphics Processing Unit: A high-
performance processor ideal for AI
workloads, simulations, and image
processing
LLM: Large Language Model: An AI model
that understands and generates human-
like text based on large datasets
API: Application Programming Interface:
A set of tools allowing different software
systems to interact and share data
IBA: Indian Bank Association: An industry
body that represents and coordinates
efforts among Indian banks
IT: Information Technology: The use of
systems and computers for processing,
storing, and transmitting information
PLI: Production-Linked Incentive: A
government scheme offering financial
rewards to boost domestic manufacturing
output
OEE: Overall Equipment Effectiveness:
A measure of how well manufacturing
equipment is utilized in terms of time,
speed, and quality
ML: Machine Learning: A form of AI
where systems learn from data to make
predictions or decisions
AGV: Automated Guided Vehicles:
Driverless vehicles used in factories to
transport materials automatically
OEM: Original Equipment Manufacturer:
A company that makes parts or products
used in another company’s end product AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 45
TSN: Time-Sensitive Networking: A
networking protocol that ensures timely
and reliable data delivery for industrial
automation
HPC Cluster: High-Performance Computing
Cluster: A network of powerful computers
used for complex calculations and AI
model training
AICTE: All India Council for Technical
Education: The regulatory authority for
technical and engineering education in
India
Biomarker: A measurable biological
indicator (molecule, gene, image, signal)
that reliably reflects a normal or pathogenic
process or drug response.
Omics is a collective term for large-scale
data layers such as genomics (DNA),
transcriptomics (RNA), proteomics
(proteins), metabolomics (metabolites)
and epigenomics.
Genomics: Study of an organism’s complete
DNA sequence.
Transcriptomics: Analysis of all RNA
transcripts in a cell or tissue
Proteomics: Large-scale study of protein
expression, structure and interaction.
Metabolomics: Comprehensive profiling of
small-molecule metabolites.
Biobank: Organised repository storing
biological samples and associated data for
research.
FAIR principles: Data management
guidelines— Findable, Accessible,
Interoperable, Re-usable.
IBDC: Indian Biological Data Centre—
national life-science data repository.
INSACOG: Indian SARS-CoV-2 Genomics
Consortium— pan-India viral sequencing
network.
ABDM: Ayushman Bharat Digital Mission—
creates India’s national health-data
infrastructure.
DALY (Disability-Adjusted Life Year): A
composite burden-of-disease metric equal
to the sum of Years of Life Lost (YLL) from
premature death and Years Lived with
Disability (YLD); one DALY represents one
healthy year of life lost.
R&D Scientists: Researchers involved
in various stages of drug development
(discovery to approval)
ANRF: Anusandhan National Research
Foundation – India’s R&D funding
body under the Ministry of Science and
Technology
BIRAC: Biotechnology Industry Research
Assistance Council, a Government of India
agency supporting biotech innovation and
startups.
NIH’s “All of US” program: A U.S. research
initiative collecting diverse health data
from over 1 million people to advance
precision medicine.
Institution-initiated trials: Clinical research
studies led and sponsored by academic
or research institutions rather than
pharmaceutical companies.
Operation Warp Speed: A U.S. government
initiative accelerating COVID-19 vaccine
development, manufacturing, and
distribution.
SEBI: Securities and Exchange Board of
India, the securities market regulator,
ensures investor protection and market
transparency.
Weighted tax deductions: Tax incentives 46 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
allowing companies to deduct multiple
of eligible R&D expenses to encourage
innovation.
Ramanujan Fellowship: Scheme offering
re-entry support to Indian scientists abroad
for research in India
Ramalingaswami Fellowship: Fellowship
program targeting postdocs in life sciences
returning to India
NAMTECH: Finishing-school style institute
model focused on deep-tech and job-
ready skills
AICTE: Apex regulatory body for technical
education in India
OCI-Fellow: Proposed track offering
incentives for Overseas Citizens of India to
return
HPC Cluster: High-performance computing
facility used for large-scale simulation and
AI research
Indian Biological Data Centre (IBDC):
India’s central omics repository, housing
genomics and bioinformatics datasets
PPP-adjusted: Purchasing Power
Parity–adjusted salary comparisons to
international levels
ICH: International Council for
Harmonisation of Technical Requirements
for Pharmaceuticals for Human Use – sets
global drug development guidelines
E2E: ICH guideline on Pharmacovigilance
Planning – ensures safety monitoring
throughout the drug lifecycle
E9: ICH guideline on Statistical Principles
for Clinical Trials – ensures consistency and
quality in trial design and data analysis
E17: ICH guideline on Multi-Regional
Clinical Trials – provides a framework for
conducting global trials efficiently
M2/M8: ICH guidelines for Electronic
Standards & Submission Format – ensures
standardized digital submissions to
regulatory authorities
CDSCO: Central Drugs Standard Control
Organization – India’s national drug
regulatory body
GCP: Good Clinical Practice – international
standard for ethical and scientific quality
in clinical trials
Data Exclusivity: A legal right preventing
generic manufacturers from referencing
an innovator’s clinical data for a defined
period
Hatch-Waxman Act: U.S. legislation that
grants data exclusivity of 5–12 years for
new drugs, protecting clinical trial data
ANRF: Anusandhan National Research
Foundation – India’s upcoming R&D
funding agency to support deep science
initiatives
CDSCO NDCT Rules (2019): India’s rules for
approval of new drugs and clinical trials;
enables accelerated pathways and ethics
compliance
DPIIT: Department for Promotion of
Industry and Internal Trade – involved in IP
reforms and innovation policy
Market Access: A Set of policies and
incentives that determine how quickly
and widely a drug reaches patients at an
affordable price
HTAIn: Health Technology Assessment
India – evaluates the value-for-money of
health technologies
HTA: Health Technology Assessment AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 47
– compares clinical efficacy, cost-
effectiveness, and equity of medical
technologies
DHR: Department of Health Research –
nodal body for HTAIn under Ministry of
Health
PM-JAY: Pradhan Mantri Jan Arogya Yojana
– India’s public health insurance scheme
covering 550M people
CDSCO: Central Drugs Standard Control
Organization – India’s national regulatory
body for drug approvals
NHA: National Health Authority –
implementing agency of PM-JAY
NRDL (China): National Reimbursement
Drug List – HTA-based system for national
drug price negotiation and reimbursement
NICE (UK): National Institute for Health
and Care Excellence – UK’s HTA body with
legal timelines post EMA approval
EMA: European Medicines Agency – central
drug approval authority for the EU
NITI Aayog: Policy think tank of the
Government of India – drives long-term
policy vision
Made-in-India Innovation Track: Proposed
track for public coverage of Indian-
developed drugs within 180 days post
CDSCO approval
DPIIT: Department for Promotion of
Industry and Internal Trade – responsible for
Startup India and pharma manufacturing
schemes
PLI: Production Linked Incentive – scheme
to boost local manufacturing with
performance-based incentives
Proprietary AI model: An AI system
exclusively owned and trained by an entity,
optimized for a specific application like
drug discovery
Foundation model: A large-scale pre-
trained AI model that can be adapted to
multiple downstream tasks
Genome-scale language model: An AI
model trained on DNA/RNA sequences to
predict the functional impact of mutations
and guide target discovery
Nucleotide Transformer: A 2.5B parameter
AI model developed by InstaDeep, NVIDIA,
and TUM that sets a new benchmark in
variant effect prediction
Single-cell foundation model: A model
trained on individual cell data (multi-omics)
to predict how drugs act across different
cell types
scGPT: A “ChatGPT for biology” AI model
developed by the Toronto Vector Institute
and UHN, trained on millions of single-cell
data points
Protein-design model: AI model that
predicts 3D protein structures and
interactions to help design new drugs
AlphaFold 3: DeepMind’s AI model that
accurately predicts 3D structures and
interactions of proteins and small molecules
Multimodal clinical-prediction model: An
AI system that integrates omics, EHR,
and phenotype data to guide clinical trial
inclusion and adverse event forecasting
Bridge2AI: U.S. NIH initiative to develop
AI-ready health datasets for training
multimodal biomedical foundation models
India Omics Commons Sandbox: Proposed
national platform to house diverse omics
datasets for AI model development
SAV (Software Assisted Vehicle): A vehicle
with embedded software capabilities, 48 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
including smart features like predictive
maintenance, OTA updates, and data
services
OTA (Over-the-Air): Wireless delivery of
software updates or patches to a vehicle’s
system without requiring service center
visits
Modular Components: Vehicle parts
designed to be independently upgraded,
replaced, or certified
Certified Components: Parts that
meet standardized safety, quality, and
performance benchmarks set by regulatory
bodies
ECU (Electronic Control Unit): Embedded
systems in vehicles that control electrical
subsystems like braking, engine, and
infotainment
ISM (India Semiconductor Mission):
A government initiative to build a
comprehensive semiconductor ecosystem
in India
28nm / 16nm: Semiconductor process
nodes indicating the transistor size used in
chip fabrication; 28nm/16nm are common
for automotive-grade chips
Wafers per Month (WPM): Unit of fab
output; indicates how many silicon wafers
can be produced each month
Brownfield Fab: A semiconductor plant
built on an existing industrial site, often
through upgrades or repurposing
Greenfield Fab: A new chip manufacturing
facility built from scratch
CHIPS and Science Act: A U.S. federal law
enacted in 2022 that allocates $52 billion
to promote domestic semiconductor
manufacturing
SMIC (Semiconductor Manufacturing
International Corporation): China’s largest
chipmaker, known for strategic investment
in 28nm and 40nm auto chips
ACMA: Automotive Component
Manufacturers Association of India; apex
body for India’s auto component suppliers
SIAM: Society of Indian Automobile
Manufacturers; works on policy and
technical issues affecting Indian OEMs
ARAI: Automotive Research Association
of India; certifies automotive components
and systems
BIS: Bureau of Indian Standards; national
standards body responsible for setting
quality norms across sectors
ISMC: International Semiconductor
Consortium, a public-private partnership
aiming to establish fabs in India
SAV: Software-Assisted Vehicle; vehicles
integrated with software-based features
like infotainment, diagnostics, and driver
assistance “Intelligent and Connected
Vehicles
Cyber-Physical System (ICV CPS):
Vehicles embedded with AI, sensors,
and communication tech to interact with
infrastructure and other vehicles in real
time.”
NIC: National Informatics Centre, India’s
technology backbone for e-governance
and digital infrastructure across
government bodies.
ISRO: Indian Space Research Organisation,
India’s national space agency responsible
for space missions, satellites, and launch
vehicles.
Zero prototype lab: A digital R&D setup
where designs are validated entirely
through simulations and AI without AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 49
building physical prototypes.
Tier 1/2 Suppliers: Automotive component
manufacturers directly (Tier 1) or indirectly
(Tier 2) supplying parts to OEMs
Validation Cycle: One complete test cycle
of a SAV feature (e.g., parking assist) under
various simulated conditions
RSU (Roadside Unit): Fixed communication
infrastructure that connects vehicles to
network and surroundings for real-time
data exchange
5G/6G Corridors: High-speed connectivity-
enabled roads to support real-time
software and V2X (vehicle- to-everything)
communication
NATRAX: National Automotive Test Tracks,
a state-of- the-art test facility in Indore,
India
STPI: Software Technology Parks of India;
organization supporting IT/ITES industries
with infrastructure and services
American Center for Mobility (ACM): A 500-
acre testing site in Michigan, USA, focused
on software- defined vehicle technologies
K-City: An 80-acre autonomous driving
testbed in South Korea, equipped for
smart mobility trials
DLS (Deep Learning Surrogates): AI
models that emulate physical simulations
to accelerate design and validation of
components
OEM (Original Equipment Manufacturer):
A company that produces parts and
equipment that may be marketed by
another manufacturer
CoE (Centre of Excellence): Specialized
institutional setups for applied research,
talent development, and innovation
AICTE: All India Council for Technical
Education – a body for curriculum reforms
and engineering standards
Micro-credentials: Short, specialized
certifications validating skillsets in niche
areas like AI in engineering
Hackathon: A competitive, time-bound
innovation challenge to solve real-world
problems
Formula-Student: An international student
competition to design and race small-
scale formula-style cars
Digital twin: A virtual representation
of a physical object or system used for
simulation, design testing, and monitoring
High-performance computing (HPC):
Large-scale computational infrastructure
used to train complex AI models and run
massive simulations
Patent grant timeline: The time taken from
patent filing to final approval and grant of
IP rights
AI-priority patent track: A fast-track
mechanism for processing patent
applications related to artificial intelligence
innovations
Model licensing scheme: A framework
enabling OEMs or Tier-1 suppliers to license
trained AI models or virtual component
validations as services
Intellectual property (IP): Legal rights
granted over inventions, datasets, models,
and processes that are original and provide
economic benefit
Ministry of Electronics and Information
Technology (MeitY): Indian ministry
overseeing digital and AI technology
advancement
Department for Promotion of Industry and 50 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
Internal Trade (DPIIT): Indian ministry
responsible for IP frameworks, industry
regulations, and startup innovation policies
PI-XAI: Physics-Informed & Explainable AI
— AI models that both obey physical laws
and provide interpretable results
CFD: Computational Fluid Dynamics —
Traditional physics-based method used
to simulate fluid flow and heat transfer in
automotive and other systems
ISO/PAS 8800:2024: International standard
for transparent and explainable AI systems
used in vehicles
ISO/IEC 42001: International standard for
ethical and accountable AI management
within organizations
AIS-140: Indian Automotive Industry
Standard related to Intelligent Transport
Systems, including GPS tracking and
emergency systems
AIS-XAI: Proposed extension of AIS
standards to include explainable AI
certifications
TÜV SÜD (Germany): Technical inspection
and certification agency offering AI
quality audits compliant with international
standards
NSCS: National Surrogate-model
Certification Sandbox — A proposed
setup at ARAI for testing and certifying AI
models using open data, physics checks,
and cybersecurity
BIS: Bureau of Indian Standards — India’s
national standards body responsible for
standard setting and certification
MHI: Ministry of Heavy Industries — Central
ministry responsible for the automotive
industry’s policy and standards governance
GST (Goods and Services Tax): A unified
indirect tax in India levied on the supply of
goods and services
Digital Patent Box: A policy regime offering
lower tax rates on profits earned from
locally developed and licensed IP
CBDT: Central Board of Direct Taxes,
responsible for direct tax policy formulation
and enforcement
PLI (Production Linked Incentive): A
government scheme offering financial
incentives based on incremental
production and sales
ICE (Internal Combustion Engine):
Traditional vehicle engines that run on
petrol or diesel
Bureau of Indian Standards (BIS): National
standards body responsible for product
certification schemes
NITI Aayog: India’s apex policy think tank
guiding economic and technological
policy reforms
Ministry of Heavy Industries (MHI): Nodal
ministry for the automotive sector, among
others
Digital twin: A virtual representation
of a physical object or system used for
simulation, design testing, and monitoring
Patent grant timeline: The time taken from
patent filing to final approval and grant of
IP rights
AI-priority patent track: A fast-track
mechanism for processing patent
applications related to artificial intelligence
innovations
Model licensing scheme: A framework
enabling OEMs or Tier-1 suppliers to license
trained AI models or virtual component
validations as services AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 51
Intellectual property (IP): Legal rights
granted over inventions, datasets, models,
and processes that are original and provide
economic benefit
Industry 4.0: The fourth industrial revolution
is marked by integration of digital
technologies like AI, IoT, and robotics into
manufacturing and industry.
Advanced Cyberinfrastructure Coordination
Ecosystem-Services & Support (ACCESS):
A U.S. initiative providing advanced
computing, data, and support services to
accelerate scientific research.
NEAT: National Educational Alliance
for Technology, a Government of India
platform to bring AI-powered learning
tools to higher education 52 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
NOTES AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 53
NOTES 54 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth
NOTES AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth | 55 56 | AI for Viksit Bharat: The Opportunity for Accelerated Economic Growth