The Ministry of Electronics & Information Technology (MeitY), under the aegis of the IndiaAI Mission, formally launched the India AI Governance Guidelines on 5 November 2025, marking a significant milestone in India’s pursuit of safe, inclusive and responsible artificial intelligence (AI) deployment across sectors. The unveiling event was attended by the country’s Principal Scientific Adviser, Ajay Kumar Sood, and MeitY Secretary S. Krishnan, alongside senior officials from the IndiaAI Mission.
The launch marks a key milestone ahead of the India-AI Impact Summit 2026, as India strengthens its leadership in responsible AI governance. The guidelines propose a robust governance framework to foster cutting-edge innovation, and safely develop and deploy AI for all while mitigating risks to individuals and society. The framework comprises four key components:
- Seven guiding principles (or “Sutras”) for ethical and responsible AI, that define India’s AI governance philosophy.
- A set of key recommendations structured along six AI governance pillars.
- An action plan with time-horizons mapped to short, medium and long-term deliverables.
- Practical guidelines directed at industry actors, developers, regulators and other stakeholders to ensure transparent, accountable deployment
Seven Guiding Principles (Sutras)
The Committee recommends that India’s AI governance framework be guided by certain principles or ‘sutras’, applicable across sectors and technologies. The first part of the document sets out seven foundational principles or sutras adapted from the RBI’s FREE-AI Committee report to guide the overall approach. These principles have been adapted for application across sectors and aligned with national priorities.
The seven guiding sutras are:
- Trust is the Foundation: Embedding trust across the entire value chain—from technology, deploying organisations to user adoption. Trust is essential to support innovation, adoption, and progress, as well as risk mitigation. Without trust, the benefits of artificial intelligence will not be realised at scale. Trust must be embedded across the value chain – i.e. in the underlying technology, the organisations building these tools, the institutions responsible for supervision, and the trust that individuals will use these tools responsibly. Therefore, trust is the foundational principle that guides all AI development and deployment in India.
- People First: Emphasising human-centric design, human oversight, and empowerment of citizens as ultimate beneficiaries. AI governance frameworks should be human-centric. That means AI systems should be designed and deployed in ways that empower individuals and reflect the value systems of the people for whom the technology is built to serve. From a governance perspective, a people-first approach means that humans should, as far as possible, have final control over AI systems, and human oversight is essential to maintain accountability. A people-first approach also prioritises human capacity development, ethical safeguards, trust and safety.
- Innovation over Restraint: Encouraging responsible innovation as a default while applying restraint only when necessary. AI-led innovation is a pathway to achieving national goals, such as socio-economic development, global competitiveness, and resilience. Therefore, AI governance frameworks should actively encourage adoption and serve as a catalyst for impactful innovation. That said, innovation should be carried out responsibly and should aim to maximise overall benefit while reducing potential harm. All other things being equal, responsible innovation should be prioritised over cautionary restraint.
- Fairness & Equity: Ensuring AI systems are inclusive, non-discriminatory, and benefit all sections of society. A key goal of India’s approach to AI governance is to promote inclusive development. Therefore, AI systems should be designed and tested to ensure that outcomes are fair, unbiased, and do not discriminate against anyone, including those from marginalised communities. AI should be leveraged to promote inclusive development while mitigating risks of exclusion, bias, and discrimination.
- Accountability: Clear allocation of responsibility among developers, deployers and other actors in the AI value chain. To ensure that India AI’s ecosystem progresses based on trust, AI developers and deployers should remain visible and accountable. Accountability should be clearly assigned based on the function performed, risk of harm, and due diligence conditions imposed. Accountability may be ensured through a variety of policy, technical and market-led mechanisms
- Understandable by Design: Requiring explainability and transparency so that end-users and regulators can understand system behaviour. Understandability is fundamental to building trust and should be a core design feature, not an afterthought. Though AI systems are probabilistic, they must have clear explanations and disclosures to help users and regulators understand how the system works, what it means for the user, and the likely outcomes intended by the entities deploying them, to the extent technically feasible.
- Safety, Resilience & Sustainability: AI systems that are secure, robust, resilient to shocks and environmentally sustainable. AI systems should be designed with safeguards to minimise risks of harm and should be robust and resilient. These systems should have capabilities to detect anomalies and provide early warnings to limit harmful outcomes. AI development efforts should be environmentally responsible and resource-efficient, and the adoption of smaller, resource-efficient ‘lightweight’ models should be encouraged.
In his remarks, Dr Sood underscored the spirit of “Do No Harm” as the core guiding ethos of the entire framework.
Six Pillars of AI Governance
Part 2 of the Guidelines sets out recommendations across six pillars grouped broadly under three domains: enablement (infrastructure, capacity building), regulation (policy & regulation, risk mitigation), and oversight (accountability, institutions). Each pillar is summarised below including key recommendations.
- Infrastructure: The Guidelines urge expansion of access to foundational compute and data resources, leveraging India’s Digital Public Infrastructure (DPI) for scale and inclusion. For instance, the document notes that over 38,231 GPUs are being made available to startups, researchers and developers at subsidised rates; a secure GPU cluster of 3,000 next-gen GPUs for strategic applications is under construction. It also highlights onboarding of 1,500 datasets and 217 AI models in the AIKosh platform from 34 entities across 20 sectors. Driving inclusive adoption, it recommends targeted incentives for MSMEs such as tax rebates, AI-linked loans and subsidised compute access.
- Capacity Building: This pillar focuses on education, training, upskilling and awareness efforts. The Guidelines call for large‐scale initiatives to deepen AI literacy especially in tier-2/3 cities, vocational institutes, and among government officials. Existing programmes such as India AI FutureSkills and FutureSkills PRIME are referenced; however, expansion is considered necessary to meet inclusive goals.
- Policy & Regulation: Here the focus is on adopting an agile, balanced regulatory framework. The Guidelines emphasise using existing laws wherever possible and making targeted amendments to address AI-specific gaps (e.g., in the IT Act 2000, DPDP Act, copyright law) rather than introducing a blanket “AI Act” immediately. For example, they point out that the IT Act needs clarity on how digital intermediaries and AI systems are classified, and how liability is apportioned when AI systems are involved.
- Risk Mitigation: The document argues for an India-specific risk assessment framework grounded in real-world evidence of harm. It recommends voluntary industry frameworks initially, supported by techno-legal solutions. In sensitive contexts (e.g., public sector, vulnerable populations), additional obligations may apply. It also addresses issues such as content authentication, deepfakes, dataset provenance, and attribution in generative AI.
- Accountability: The Guidelines propose a graded liability regime: responsibility should be allocated based on the function performed, level of risk and due diligence observed. Transparency in the AI value chain—developers, deployers, third-party model providers—is emphasised, so that regulators can design proportionate and effective oversight.
- Institutions: A “whole-of-government” institutional architecture is recommended. Specifically, the creation of an AI Governance Group (AIGG) is proposed, supported by a Technology & Policy Expert Committee (TPEC) and backed by a technical body, the AI Safety Institute (AISI). Sectoral regulators, standard-setting bodies (e.g., BIS, TEC) and state governments are expected to participate.
Action Plan: Short, Medium, Long-Term
The Guidelines map out an Action Plan with clear time horizons.
- Short-term: Establish key governance institutions; develop India-specific risk frameworks; adopt voluntary commitments from industry; suggest legal amendments; develop clear liability regimes; expand infrastructure access; launch awareness campaigns; increase access to AI safety tools.
- Medium-term: Publish common standards; amend relevant laws/regulations; operationalise AI incident reporting systems; pilot regulatory sandboxes; expand integration of AI with DPI; continue capacity building and standard-setting.
- Long-term: Review and update governance frameworks periodically for sustainability; draft new laws based on emerging risks and capabilities; ensure the digital ecosystem remains resilient and adaptive.
Practical Guidelines for Stakeholders
Part 4 of the document offers pragmatic guidance:
- For Industry / Developers / Deployers: Comply with existing Indian laws; adopt voluntary principles/standards; publish transparency or “self-assessment” reports; implement grievance redressal mechanisms; adopt techno-legal risk-mitigation methods.
- For Regulators: Support innovation while proactively addressing identified harms; avoid unduly heavy compliance burdens; promote techno-legal approaches; adopt flexible and periodically reviewed frameworks to keep pace with evolving technology.
In unveiling the guidelines, Secretary S. Krishnan emphasised that the focus remains on leveraging existing legislation wherever possible, with human-centricity at the heart of AI’s deployment to serve citizens’ lives while addressing harms. The timing of the announcement is significant, ahead of the planned India AI Impact Summit 2026 in New Delhi.
