IndiaAI Mission — The Regulatory Architecture Explained

The IndiaAI Mission is the Government of India's Rs. 10,371.92 crore, five-year executive scheme for artificial intelligence, approved by the Union Cabinet on 7 March 2024 and delivered through the IndiaAI Independent Business Division of Digital India Corporation under the Ministry of Electronics and Information Technology. Its seven pillars — Compute Capacity, Innovation Centre, Datasets Platform, Application Development Initiative, FutureSkills, Startup Financing and Safe & Trusted AI — together constitute India's primary AI industrial-policy architecture and the funding frontier for Indian AI founders.

TL;DR for founders

IndiaAI Mission is how the Government of India will spend Rs. 10,371.92 crore on AI over five years. If you want subsidised GPU compute, dataset access, sector-application grants (healthcare, agriculture, education), deep-tech startup capital, FutureSkills funding or Safe-and-Trusted-AI research support, this is the door. Apply through calls for proposals on indiaai.gov.in — run by the IndiaAI Independent Business Division under Digital India Corporation, under MeitY. The Mission is an executive scheme, not a statute — so it does not create new AI offences. But taking Mission money layers grant-specific strings (IP flow-back to the Government of India, audit rights, publication requirements, open-access for Indian public-sector users) on top of your existing DPDP Act, IT Rules 2021 (including the 20 February 2026 Synthetic Media Amendment), CERT-In Directions 2022 and sectoral compliance stack.

Why India launched its own AI mission

Between late 2022 and early 2024 three policy currents converged. First, the United States Executive Order 14110 (30 October 2023) reshaped the global discussion on AI safety testing; second, the European Union reached political agreement on the Artificial Intelligence Act (text adopted March 2024); and third, India's own Digital India Dialogues surfaced acute industry demand for subsidised compute and access to clean, Indian-language datasets. The Union Cabinet's 7 March 2024 decision responded by creating a single whole-of-government AI programme — an industrial-policy vehicle that would (a) underwrite the compute and data scarcity preventing Indian AI innovation, (b) park a governance research function inside the Mission rather than in a regulator, and (c) preserve MeitY's statutory rule-making authority under the Information Technology Act, 2000 ('IT Act') for binding AI obligations. The Mission is therefore best read as industrial policy with regulatory scaffolding, not as a standalone AI statute.

India has not yet enacted a horizontal AI law comparable to the EU AI Act. The Digital India Act, which MeitY has signalled will eventually consolidate and replace the IT Act and the IT Rules 2021, has not yet been introduced in Parliament as at April 2026. In the interim, the Mission's Safe & Trusted AI pillar — and the IndiaAI Safety Institute announced by the Minister for Electronics and Information Technology on 30 January 2025 — carries the substantive governance workstream, while binding obligations continue to flow from the existing statutory chain.

Governance architecture — MeitY to DIC to IndiaAI IBD

The Mission's delivery chain has three layers.

1. Ministry of Electronics and Information Technology ('MeitY'). MeitY is the administering ministry, the rule-making authority under Section 87 IT Act, and the convenor of all AI-related advisories (including the MeitY Advisory dated 1 March 2024 as revised on 15 March 2024). MeitY owns policy direction, inter-ministerial coordination and statutory notification authority.
2. Digital India Corporation ('DIC'). DIC is a Section 8 not-for-profit company under the administrative control of MeitY. It is the Mission's fiscal and operational vehicle — it signs grant agreements, disburses funds, operates public-facing platforms and houses the IndiaAI Independent Business Division.
3. IndiaAI Independent Business Division ('IndiaAI IBD'). The IBD is the Mission secretariat inside DIC. It issues calls for proposals, evaluates applications, disburses grants, operates indiaai.gov.in, and coordinates with line ministries (Health & Family Welfare, Agriculture & Farmers Welfare, Education). The IBD is the direct counter-party for any founder contracting with the Mission.

This three-layer structure matters legally because a grant agreement is signed with DIC / IndiaAI IBD — a corporate entity — and not with MeitY directly. Disputes are therefore contractual in the first instance; writ jurisdiction under Article 226 of the Constitution is available where state action is established, but the baseline remains contract law under the Indian Contract Act, 1872.

The seven pillars — specifics and figures

The Cabinet approval structures the Mission across seven pillars. The figures below are drawn from PIB Press Release PRID 2012355 and the indiaai.gov.in portal.

1. IndiaAI Compute Capacity

Establishment of a scalable AI compute ecosystem comprising more than 10,000 Graphics Processing Units built through public-private partnership. The capacity is intended to service startups, researchers, academia and Government users. Access is rationed through an IBD-operated empanelled-user framework with tiered pricing. Calls for expression of interest are routed through indiaai.gov.in, with compute allocations granted in GPU-hour bundles.

2. IndiaAI Innovation Centre

Development and deployment of indigenous Large Multimodal Models ('LMMs') and domain-specific foundational models in critical sectors. The Innovation Centre operates on a grant-cum-partnership model with Indian research teams; the IBD retains a Government-use licence for Mission-wide deployment. Priority is expressly given to Indian languages and Indian-context datasets.

3. IndiaAI Datasets Platform

A unified, one-stop platform for access to quality non-personal datasets for AI innovation. The platform is scoped to non-personal data to avoid collision with the Digital Personal Data Protection Act, 2023 ('DPDP Act'). Where a founder's downstream use re-identifies individuals, Section 4 and Section 6 DPDP Act obligations apply in full — the platform does not operate as a Section 7 DPDP Act legitimate-use ground or a Section 17 DPDP Act exemption.

4. IndiaAI Application Development Initiative

Grants and deployment support for production-grade AI products in priority sectors — healthcare, agriculture, education and governance — where the IBD issues problem-statement-specific calls for proposals. Typical grant sizes run from seed (Rs. 1–2 crore) to growth (Rs. 10 crore and above) with milestone-linked disbursement. This pillar is the Mission's most founder-facing track.

5. IndiaAI FutureSkills

Curriculum support, Data and AI Labs in Tier 2 and Tier 3 cities, and expansion of AI courses at undergraduate, master's and doctoral level. Delivered in partnership with academic institutions and under the Ministry of Education framework.

6. IndiaAI Startup Financing

A streamlined access-to-funding track for deep-tech AI startups, operated in conjunction with Government venture instruments (including SIDBI's Fund of Funds for Startups) and private co-investors. DPIIT recognition under Startup India is typically a prerequisite. Grants and equity-linked instruments are both available depending on the specific call for proposals.

7. Safe and Trusted AI

The regulatory-signalling pillar. Funds Responsible-AI projects, indigenous testing and evaluation tooling, self-assessment checklists, and governance-framework research. Houses the IndiaAI Safety Institute ('AISI') announced on 30 January 2025, which operates on a Hub-and-Spoke model engaging academia, startups, industry and line ministries to produce AI-safety research, evaluation methodologies and Responsible-AI reference frameworks calibrated to Indian social, economic, cultural and linguistic diversity. AISI is a research-and-standards body — it is not the AI regulator; binding rules continue to emanate from MeitY under the IT Act and from sectoral regulators.

Procurement and grant pathways for Indian startups

Indian AI founders engage the Mission through three principal pathways. Each has a distinct application lifecycle, evaluation rubric and contractual posture.

1. Call for proposals ('CFP') track — grants. The IBD issues thematic CFPs on indiaai.gov.in — for example, specific healthcare-AI problem statements, specific Indian-language LMM targets, or specific Safe-and-Trusted-AI research deliverables. Applicants file online; the IBD runs a two-stage evaluation (expression of interest, then full proposal) with a multi-member Technical Evaluation Committee. Selected applicants sign a grant agreement with DIC that includes milestone, IP, reporting and audit clauses.
2. Compute access track — empanelment. Founders seeking subsidised GPU compute apply for empanelment under the Compute Capacity pillar. Empanelment grants access to tiered GPU-hour bundles at subsidised rates. Empanelment agreements include fair-use, no-military-use and no-model-fine-tune-for-third-party clauses; breach typically triggers compute-access revocation rather than statutory penalty.
3. Dataset access track — portal registration. Founders requiring dataset access register on the IndiaAI Datasets Platform and accept a click-through data-use licence. The licence includes a non-re-identification warranty and a DPDP-compliance covenant running from the user to the IBD. Breach exposes the user to contractual remedies and, where personal data is involved, to DPDP Act penalties up to Rs. 250 crore per instance under the Schedule to the DPDP Act.

Safe & Trusted AI — the regulatory sub-pillar

The Safe and Trusted AI pillar is the Mission's policy bridge between industrial funding and binding AI law. Its three operative outputs are:

1. Research on evaluation methodologies — benchmark design, red-team protocols, bias testing frameworks, Indian-language reliability testing, election-integrity AI testing, and deepfake-detection reference implementations.
2. Self-assessment checklists and governance templates — reference documents that Mission-funded (and non-Mission) AI deployers can adopt to demonstrate due diligence. Prudent practice is to treat these checklists as the operational expression of "due diligence" under Section 79(2)(c) of the IT Act, even before they become binding by rule amendment.
3. Stakeholder convening — structured engagement with line ministries, startups, industry associations and academia, with outputs feeding into subsequent MeitY rule-making cycles.

The Safe and Trusted AI pillar connects directly to (a) the MeitY Advisory dated 1 March 2024 on labelling, consent pop-ups and originator identification; (b) the Synthetic Media Amendment Rules, 2026 (G.S.R. 120(E), notified 10 February 2026, effective 20 February 2026) which introduced the three-hour takedown window for a broad set of illegal content, a two-hour window for non-consensual intimate imagery and child sexual abuse material, and binding synthetic-media labelling under Rule 3(3) of the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 ('IT Rules 2021'); and (c) the DPDP Act consent regime where personal data trains the model.

Interaction with sectoral AI regulation

Mission-funded AI products operating in regulated sectors layer an additional regulator on top of the MeitY baseline.

Mission funding does not override sectoral rule-making. A Mission-funded healthcare AI startup remains obliged to obtain CDSCO clearance before commercial deployment; a Mission-funded fintech AI startup remains bound by RBI digital-lending rules irrespective of Mission support.

Practitioner analysis

1. Structuring an AI startup to maximise IndiaAI eligibility

Counsel advising a founder team on AI-startup incorporation should layer four incorporation decisions. First, incorporate as a private limited company under the Companies Act, 2013 — Limited Liability Partnerships are generally ineligible for deep-tech grant pathways and sector-specific CFPs. Second, complete DPIIT recognition under Startup India early; many IBD CFPs require DPIIT number on application. Third, confirm Indian tax residency of the company (directors meeting in India, board and management control in India); this matters both for Mission eligibility and for Section 80-IAC Income-tax Act, 1961 benefits. Fourth, if there is a foreign holding company, map the funds-flow structure to avoid triggering FEMA (Foreign Exchange Management Act, 1999) non-debt-instrument rule complications — Mission grants disbursed to an Indian subsidiary that then up-streams IP to a foreign parent can generate transfer-pricing, FEMA and grant-breach exposure simultaneously.

2. IP and data ownership in Mission-funded research

Grant agreements issued under the Innovation Centre and the Application Development Initiative typically reserve three Government-favouring IP rights. Counsel should pressure-test each clause against pre-existing investor rights.

1. Foreground IP — the IP developed under the grant — is generally owned by the recipient, subject to a non-exclusive, royalty-free Government-use licence for Indian public-sector deployment. Benchmark against investor SHA clauses on "all IP exclusively owned by the company" to avoid silent breach.
2. Background IP — pre-existing recipient IP used in the grant — is typically subject to a non-assertion undertaking during the grant window. Benchmark against customer master services agreements that warrant non-encumbered IP delivery.
3. Audit rights — the IBD typically obtains source-code and model-artefact audit rights during the grant window and for a trailing period post-completion. This interacts with customer NDAs and investor reporting rights; a waterfall of confidentiality sub-permissions is usually required.

3. Compliance stack for Safe & Trusted AI participants

A Mission participant under the Safe & Trusted AI pillar should maintain a live compliance register capturing: (i) bias and reliability test logs for each production model with dated entries; (ii) red-team test results mapped to Rule 3(1)(b) IT Rules 2021 content categories; (iii) model cards with training-data provenance and version hashes; (iv) Rule 3(3) IT Rules 2021 synthetic-media labelling and watermark implementation evidence; (v) Rule 3(1)(d) three-hour takedown SOPs operational by 20 February 2026; (vi) DPDP Act consent-collection records under Section 6 DPDP Act and, where applicable, Data Protection Impact Assessments under Section 10 DPDP Act; and (vii) CERT-In Directions dated 28 April 2022 — six-hour incident reporting SOP and 180-day log-retention under Direction (v). This register operationalises "due diligence" under Section 79(2)(c) IT Act and is the most likely evidentiary battlefield if safe-harbour is challenged in litigation.

Founder checklist

Founder checklist

• Before applying — confirm DPIIT Startup India recognition, Indian incorporation, and (if applicable) FEMA-compliant funds-flow structure.
• CFP selection — read each specific call for proposals on indiaai.gov.in in full; eligibility, IP terms and audit rights vary materially across pillars and CFPs.
• Grant negotiation — pressure-test foreground-IP, audit-rights and public-sector-use clauses against existing investor SHA and customer MSAs; seek targeted carve-outs rather than blanket acceptance.
• Operational compliance — maintain a dated bias / reliability test register, Rule 3(3) synthetic-media labelling evidence, Rule 3(1)(d) three-hour takedown SOP (operational from 20 February 2026), DPDP Act consent records and CERT-In 2022 six-hour incident reporting SOP from day one.
• Post-award — track IndiaAI Safety Institute publications and fold their self-assessment checklists into internal audit protocols; bank evidence for the Digital India Act transition.

Frequently asked questions

Can a foreign-incorporated company apply for IndiaAI Mission grants? The IndiaAI Application Development Initiative and the IndiaAI Startup Financing pillar are principally directed at Indian-incorporated deep-tech AI companies, typically DPIIT-recognised startups under the Startup India initiative. Foreign-incorporated companies cannot access those two pillars directly. The IndiaAI Compute Capacity empanelled-user framework does admit research collaborations with foreign academic institutions where an Indian nodal institution is the lead applicant. Each call for proposals routed through indiaai.gov.in specifies its own eligibility gate — Indian tax residency and incorporation are typically mandatory for grant-linked pillars.

What intellectual property do we retain when we accept an IndiaAI grant under the Innovation Centre or Application Development Initiative? Grant agreements issued by the IndiaAI Independent Business Division under Digital India Corporation typically reserve three government-favouring IP rights: a royalty-free Government-use licence to the foreground IP for Indian public-sector deployment, a non-assertion undertaking on background IP during the grant window, and audit rights over source code and model artefacts. Residual commercial IP remains with the grant recipient. Actual clause language differs by call for proposals — founders should benchmark each call's IP schedule against their existing investor shareholders' agreement and customer licensing terms before signing.

What is the Safe and Trusted AI pillar's compliance scope? The Safe and Trusted AI pillar funds indigenous Responsible-AI tooling, self-assessment checklists and governance frameworks delivered through the IndiaAI Safety Institute announced on 30 January 2025. Grant recipients under this pillar typically agree to open publication of safety research, deposit of evaluation datasets to the IndiaAI Datasets Platform, and participation in red-team exercises convened by the Institute. The pillar is research-and-standards-led — it does not carry statutory offences; binding obligations on AI output continue to flow from Rule 3(1)(b), Rule 3(1)(d) and Rule 3(3) of the IT Rules 2021 and from the MeitY Advisory dated 1 March 2024.

Does the IndiaAI Datasets Platform interact with DPDP Act consent requirements? Yes. The platform is scoped to non-personal datasets by design, but the moment a dataset combines with personal data — for example, by linking anonymised clinical records to outcomes that become re-identifiable — the combined dataset re-enters Section 2(t) of the Digital Personal Data Protection Act, 2023 and the full Data Fiduciary regime under Section 4, Section 6 and Section 8 DPDP Act applies. Mission participation is not a Section 7 DPDP Act legitimate-use ground and does not disapply consent obligations; founders should maintain a documented decision tree distinguishing non-personal, pseudonymous and personal data flows.

How does IndiaAI Mission participation interact with Startup India DPIIT recognition? Startup India recognition under the Department for Promotion of Industry and Internal Trade ('DPIIT') is typically a prerequisite for the IndiaAI Startup Financing pillar and several Application Development Initiative calls. Recognition delivers tax benefits under Section 80-IAC of the Income-tax Act, 1961, patent-fee concessions, self-certification under labour and environment laws, and access to the Fund of Funds for Startups operated by the Small Industries Development Bank of India. IndiaAI layers sector-specific AI funding on top of these Startup India benefits — it does not duplicate or replace them. Founders should confirm DPIIT status first and then map the Mission grant into the existing Startup India compliance calendar.

Sources

• PIB — Cabinet Approves Ambitious IndiaAI Mission (PRID 2012355, 7 March 2024)
• IndiaAI Mission portal
• IndiaAI — Cabinet approves India AI Mission at Rs 10,372 crore
• Prime Minister of India — Cabinet IndiaAI Mission release
• Principal Scientific Adviser — AI Mission
• MeitY — Press Releases and Advisories
• Digital Personal Data Protection Act, 2023 — India Code
• Information Technology Act, 2000 — India Code
• Startup India — DPIIT Recognition

This explainer is part of Veritect's Digital, Data & AI Law vertical. It is an original analysis prepared exclusively from Tier 1 government sources — the Press Information Bureau, the Prime Minister's Office, the Principal Scientific Adviser, MeitY, the IndiaAI portal operated by Digital India Corporation's IndiaAI Independent Business Division, India Code, and Startup India — and does not reproduce or paraphrase any third-party commentary.