India AI Governance Guidelines (MeitY, 2025)
Issued by
Ministry of Electronics and Information Technology (MeitY), Government of India
- September 30, 2026 · Correction — Re-identified the instrument as the November 2025 India AI Governance Guidelines, removed the withdrawn approval requirement, and attributed labelling to the 2026 IT Rules (fact-check finding). (Cody Maxwell)
- October 1, 2026 · Correction — Rewrote the practical steps, newsletter hook, search description, and audience fields to match the entry's corrected content. (Cody Maxwell)
MeitY released the India AI Governance Guidelines on 5 November 2025 under the IndiaAI Mission. They are voluntary principles and recommendations that favour self-regulation over approvals or licensing. Binding labelling of AI-generated content comes separately from the 2026 amendments to India's IT Rules.
Applies To
Overview
MeitY released the India AI Governance Guidelines on 5 November 2025 under the IndiaAI Mission. A drafting committee chaired by Professor Balaraman Ravindran of IIT Madras, set up in July 2025, wrote them after an earlier draft drew more than 2,500 comments. The guidelines set out seven principles, or sutras, with recommendations, an action plan, and practical guidance. They propose new institutions, including an AI Governance Group and a technology and policy expert committee, alongside the existing AI Safety Institute. They favour voluntary measures and self-certification, and advise against compliance-heavy rules such as mandatory approvals or licensing unless strictly necessary. They are separate from MeitY's 2024 advisories. A 1 March 2024 advisory required government permission before deploying under-tested AI models, but a revised advisory on 15 March 2024 dropped that requirement. Binding duties now come from other instruments. The IT (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026 took effect on 20 February 2026. They require prominent labels and permanent provenance metadata on synthetically generated content. The Digital Personal Data Protection Act, 2023 also applies to AI that processes personal data.
Key Requirements
- •The guidelines are voluntary and favour self-regulation and self-certification.
- •They advise against mandatory approvals or licensing for AI unless strictly necessary.
- •MeitY's 2024 requirement for government permission before deploying under-tested models was withdrawn on 15 March 2024.
- •Since 20 February 2026, the amended IT Rules require labels and provenance metadata on synthetically generated content.
- •AI that processes personal data must comply with the Digital Personal Data Protection Act, 2023.
- •Intermediaries must keep grievance mechanisms under the IT Rules.
What Your Organization Must Do
- →Treat the MeitY guidelines as voluntary principles, not as binding duties on your AI systems.
- →Check whether your platform publishes synthetically generated content, since the amended IT Rules require labels and provenance metadata from 20 February 2026.
- →Map your AI data pipelines against the Digital Personal Data Protection Act, 2023 wherever personal data is processed.
- →Remove any internal rule that assumes MeitY permission is needed before deploying under-tested models, since that requirement was dropped on 15 March 2024.
- →Review grievance mechanisms required of intermediaries under the IT Rules, because breaches can cost safe harbour protection.
- →Consider voluntary self-certification against the seven principles, and record the reasoning for internal governance evidence.
Playbook Guidance
Step-by-step implementation guidance for compliance teams.
Frequently Asked Questions
- Does the India AI Governance Framework have the force of law?
- No, the guidelines are voluntary. Binding duties come from the IT Act, 2000, the IT Rules as amended in 2026, and the Digital Personal Data Protection Act, 2023.
- Which companies need MeitY approval before deploying AI models in India?
- None currently: MeitY's 1 March 2024 advisory required permission for under-tested models, but the 15 March 2024 revision dropped it, and the 2025 guidelines advise against such approvals.
- How does the India AIGF interact with the Digital Personal Data Protection Act 2023?
- The frameworks overlap wherever AI systems process personal data of Indian residents. Enterprises must satisfy DPDPA obligations, including lawful basis, consent management, and data principal rights, alongside AIGF requirements. A gap analysis mapping AI data pipelines against DPDPA is a practical first compliance step.
- What are the traceability obligations under the India AI Governance Framework?
- The guidelines set no traceability duty, but the 2026 IT Rules require provenance metadata on synthetic content. The DPDP Act imposes no general data localization; it allows transfers abroad unless the government restricts a country.
- Does the India AIGF impose content labeling requirements for deepfakes?
- The guidelines do not, but the IT Rules amendments in force from 20 February 2026 do. Intermediaries must label synthetically generated content prominently and attach permanent provenance metadata.
- What penalties apply if a company ignores MeitY advisories under the India AIGF?
- The guidelines carry no penalties. Breaching the IT Rules can cost intermediaries their safe harbor protection and lead to takedown or blocking orders.