
India AI regulation is set for a major step forward, with IT Minister Ashwini Vaishnaw saying the government will issue a consultation paper on artificial intelligence within a month, focused on safety, deepfakes and a human-first approach.
The proposed framework is expected to examine safety, deepfakes, psychological effects, liability and the degree of autonomy permitted for agents, including whether they may retain and reuse data, according to the Indian Express reporting cited in an AI briefing. The minister called safety related to AI use the most important point.
The remarks were reported by DD News, the government’s public broadcaster, from an event on October 8, 2026, according to reports. As of October 13, no paper had been published, with no title, comment deadline or submission address announced.
A change of position
The move marks a change from the government’s earlier position that existing laws were sufficient, according to the briefing. India is a major technology market and service-delivery location, so a broader regime could affect developers, outsourcing providers and international organisations that process Indian data or deploy customer-facing systems there.
Rules on AI-made content already apply: the IT Rules were amended on February 10, 2026, to cover deepfakes and other synthetically generated information, according to reports. The Press Information Bureau’s record from December 2025 says a new horizontal AI law is not required at this stage, a position the minister’s remarks did not address.
What to watch
Policy teams will need to compare proposed liability and consent rules with UK and EU requirements, according to the briefing, and procurement should preserve the ability to restrict agent autonomy, data retention and model changes by jurisdiction.
The minister’s spoken estimate points to about November 8, 2026, for the paper’s release. Until then, the framework remains promised rather than published, and organisations operating in India are watching for the consultation’s scope: whether it becomes the base for binding regulation, and how far it goes on deepfakes, agent autonomy and accountability for AI harms.
Until the paper appears, businesses face a waiting game: existing IT Rules already cover deepfakes and synthetic content, but the broader questions of liability, agent autonomy and safety standards remain unanswered. The consultation, when it lands, will be the first real indication of how far New Delhi intends to go.
Related reading: Canada Launches National AI Council to Guide AI for All Strategy; see also Jaishankar Warns Terrorism as Statecraft Can Never Be Justified.

