AI Regulation in India: What the Coming Consultation Paper Means

India is about to write its first real rulebook for artificial intelligence. On 8 October, IT Minister Ashwini Vaishnaw said the “time has come” for AI regulation in India and that MeitY will publish a regulation-focused consultation paper within a month, with safety and deepfakes at the top of the list. For startups, developers and platforms that have built on India’s light-touch approach so far, that is a significant change of direction.

Here is what was announced, how the government’s thinking has shifted, and what you should prepare for.

What Vaishnaw Announced

Speaking at the release of the World Bank’s World Development Report, Vaishnaw said the paper “will form the base for the regulation which has to come.” The government will move ahead with rules based on the feedback it receives.

The paper is expected to focus on:

• AI safety: which the minister called the most important consideration.
• Deepfakes: described as a “nuisance” whose quality is improving every day.
• User harm: including psychological harm linked to AI tools.
• A human-first approach and skilling: so that no section of society is left behind.

Vaishnaw also said India needs a “techno-legal” solution rather than law alone, and that industry must carry the main responsibility for cybersecurity, safety and deepfakes.

Source: ThePrint (PTI), Outlook Business

From “No AI Law” to a Dedicated Framework

This is a clear policy shift. In April 2023, Vaishnaw told Parliament the government was not considering a law to regulate AI. The India AI Governance Guidelines, released by MeitY in November 2025, also avoided a standalone law and leaned on existing ones such as the IT Act and the DPDP Act.

The tone began changing in 2026. In June, Vaishnaw said a new law was needed because the IT Act dates from 2000, before modern AI. In July, MeitY Secretary S. Krishnan said stakeholder talks would be followed by drafting a separate AI law. The consultation paper is the first formal step on that path.

Source: MediaNama, Mondaq

Deepfake Rules Already in Force

India is not starting from zero on deepfakes. Amendments to the IT Rules, notified on 10 February 2026 and effective from 20 February, already require platforms to:

• Prominently label permitted AI-generated (synthetic) content.
• Attach permanent metadata or provenance markers that users cannot strip out.
• Act on government or court takedown orders within three hours, down from 36 hours.

Large platforms with over 50 lakh Indian users must also ask users to declare synthetic content and verify those declarations. The new paper is likely to build on these duties rather than replace them.

Source: Hogan Lovells

What AI Regulation in India Could Mean for Builders

According to Business Today, the IT Ministry is exploring a consent-based framework for synthetic content, possible limits on how much autonomy agentic AI systems can have, and regulatory sandboxes for high-risk uses. None of this is final, but it signals where compliance work may land.

• Developers: build labelling, provenance and audit logs into AI features now, not after the rules arrive.
• Founders: teams shipping AI agents that act on a user’s behalf should document safeguards and human checkpoints.
• Businesses: review where you use generative AI in marketing or customer service, since primary responsibility is likely to sit with the deploying company.
• Users: expect clearer labels on AI content and faster removal of harmful deepfakes.

Source: Business Today

The Takeaway

The details of AI regulation in India will only be clear once MeitY publishes the paper, expected around early November. When it does, read it and respond. A consultation is the one stage where startups, developers and users can shape the rules before they become obligations.

Note: This article is based on public statements and reports as of 16 October 2026. It does not constitute legal advice.