India AI Governance Guidelines explained in one line: MeitY’s framework, released on 5 November 2025 under the IndiaAI Mission, decides that India does not need a separate AI law for now and that existing laws on IT, data protection, consumer protection, copyright and sector regulation should apply to AI, with targeted amendments where gaps appear. The Guidelines set out seven sutras (principles), six pillars of recommendations, new institutions such as the AI Governance Group, and practical expectations for companies: comply with existing law, publish transparency reports, offer grievance redressal and keep human oversight.
| Point | Position |
|---|---|
| Issued by | Ministry of Electronics and Information Technology (MeitY), under the IndiaAI Mission |
| Date | 5 November 2025, after a public consultation on a committee draft |
| Legal nature | Policy framework, not a statute: no offences, no penalties of its own |
| Central choice | No standalone AI law at this stage; apply existing laws and amend where needed |
| Structure | Seven sutras, six pillars, an action plan, practical guidelines for industry and regulators |
| New bodies | AI Governance Group (AIGG), Technology and Policy Expert Committee (TPEC), a strengthened AI Safety Institute |
| Launch theme | “Do No Harm” |
India AI Governance Guidelines explained: the seven sutras
| Sutra | What it asks of an organisation |
|---|---|
| Trust is the foundation | Build systems people can rely on; without trust, adoption fails |
| People first | Human oversight and control at critical points; design for the people affected |
| Innovation over restraint | Prefer responsible innovation to precautionary bans |
| Fairness and equity | Test for bias and avoid discrimination, including against marginalised groups |
| Accountability | Clear responsibility across the AI value chain, in proportion to role and risk |
| Understandable by design | Disclosures and explanations users and regulators can follow |
| Safety, resilience and sustainability | Robust, secure systems, with attention to environmental cost |
The same seven sutras appear in the Reserve Bank of India’s FREE-AI committee report of 13 August 2025, so a bank or NBFC meets the same vocabulary from two regulators.
The six pillars and the new institutions
The recommendations sit in six pillars, grouped into three domains:
- Enablement: infrastructure (compute, data, digital public infrastructure) and capacity building (skills and awareness).
- Regulation: policy and regulation (reviewing and amending existing laws) and risk mitigation (an India-specific risk framework and techno-legal tools).
- Oversight: accountability (graded responsibility, transparency, grievance redressal) and institutions.
| Body | Role |
|---|---|
| AI Governance Group (AIGG) | Permanent inter-agency body coordinating AI policy across ministries and regulators |
| Technology and Policy Expert Committee (TPEC) | Expert advice to the AIGG on technology, risks, global developments and legal changes |
| AI Safety Institute (AISI) | Research, risk assessment, safety testing and standards |
| Sector regulators | RBI, SEBI, IRDAI and others continue to regulate AI in their sectors |
The action plan runs in three stages: short term (set up the AIGG and TPEC, risk frameworks, voluntary commitments, awareness), medium term (standards, AI incident reporting systems, legal amendments, regulatory sandboxes) and long term (review and new law for new risks).
Learn this properly, with a certificate. Module 1 of this course walks through MeitY’s Guidelines, the sutras, pillars and institutions, and Modules 5 and 6 turn them into an AI inventory, risk register, model policy and 90-day plan. AI Law and Governance in India, ₹599: self-paced lessons, module quizzes and a 30-question final test; pass at 60% for a verifiable Lexfolio certificate.
Who must do what under the Guidelines
The Guidelines do not bind anyone by themselves. But their practical expectations for developers and deployers describe what due care will look like when a regulator, client or court asks questions.
- Comply with existing Indian law: the DPDP Act, 2023 and Rules, 2025; the IT Act, 2000 and IT Rules (including the February 2026 synthetic content amendment); consumer, copyright and sector rules.
- Adopt voluntary frameworks and standards for responsible AI, and commit publicly where sensible.
- Publish transparency reports on AI use and risk management; sensitive detail can go to regulators confidentially.
- Provide grievance redressal so a person affected by an AI output or decision can complain and get an answer.
- Use techno-legal safeguards: watermarking and provenance, bias detection, privacy-preserving techniques.
- Accept responsibility in proportion to role and risk: developers for the model, deployers for how they use it.
- Keep humans at critical decision points and test fairness with representative data.
On copyright, the drafting committee suggested considering a text and data mining exception. A separate DPIIT committee later proposed a different, licensing-based model in December 2025, which is still only a proposal.
A worked example: what a mid-sized company does next
Arjun heads compliance at an insurance broker in Gurugram. His CEO asks: “If India has no AI law, do we need to do anything?” Arjun’s answer, built on the Guidelines:
- List every AI tool in use, including AI features switched on inside existing software.
- Rate each use by who is harmed if it is wrong; treat claims and underwriting decisions as high risk.
- Adopt a short AI policy and an approved tools list.
- Open a grievance channel for customers affected by automated decisions.
- Check IRDAI’s rules on fair treatment, outsourcing and cyber security, which already apply.
Ninety days of this work is the evidence that the company governed its AI with care.
Frequently asked questions
Can you give the India AI Governance Guidelines explained simply?
They are MeitY’s 5 November 2025 policy framework saying India will govern AI through existing laws rather than a new AI statute for now. They set seven principles called sutras, six pillars of recommendations and new coordinating bodies, and expect companies to be transparent, offer grievance redressal and keep human oversight.
Are the India AI Governance Guidelines legally binding?
No. They are a policy framework without penalties of their own. But the laws they point to, such as the DPDP Act, the IT Act and Rules, consumer law and sector regulations, are binding and already apply to AI.
What are the seven sutras of AI governance in India?
Trust is the foundation; people first; innovation over restraint; fairness and equity; accountability; understandable by design; and safety, resilience and sustainability. RBI’s FREE-AI report of August 2025 uses the same seven.
What is the AI Governance Group?
It is the permanent inter-agency body the Guidelines propose to coordinate AI policy across ministries and regulators. It is advised by the Technology and Policy Expert Committee, while the AI Safety Institute handles technical testing and standards.
Quick quiz
Q1. When did MeitY release the India AI Governance Guidelines?
- 11 August 2023
- 5 November 2025
- 13 November 2025
- 20 February 2026
Show answer
Answer: B. 5 November 2025, under the IndiaAI Mission.
Q2. The Guidelines’ central choice on law is:
- A new AI Act with licences
- A ban on generative AI
- Apply existing laws, with targeted amendments
- Leave AI to courts alone
Show answer
Answer: C. No standalone AI law at this stage.
Q3. Which is NOT one of the seven sutras?
- People first
- Accountability
- Profit before safety
- Understandable by design
Show answer
Answer: C. The seventh sutra is safety, resilience and sustainability.
Turn the Guidelines into a working programme with AI Law and Governance in India, ₹599. For everyday staff use, see AI at Work: Using ChatGPT and AI Tools Safely, Smartly and Legally, ₹299.
Sources
- Ministry of Electronics and Information Technology, India AI Governance Guidelines, 5 November 2025
- Reserve Bank of India, Report of the Committee on FREE-AI, 13 August 2025
- DPIIT committee, Working Paper on Generative AI and Copyright, Part 1, 8 December 2025
- Digital Personal Data Protection Act, 2023; DPDP Rules, 2025; Information Technology Act, 2000 and IT Rules, 2021 as amended
This page is for education and general information, not legal advice for a specific case. Law as on 31 October 2026.
Useful books for this topic
- Bare acts of the new criminal laws (BNS, BNSS, BSA)
- Code of Civil Procedure bare act
- Legal drafting and conveyancing books
- Constitution of India (coat pocket edition)
As an Amazon Associate, Lexfolio earns from qualifying purchases. Prices and availability are shown on Amazon.
Leave a Reply