India Will Regulate AI: Consultation Paper Due in a Month
3 min readIndia is moving from guidelines to law on artificial intelligence. Union IT minister Ashwini Vaishnaw said on October 8 that the Ministry of Electronics and IT will publish a consultation paper on India AI regulation within a month, covering AI safety, deepfakes, user harm, a human-first approach and skilling. “The time has come for AI regulation,” he said.
Background
Until now New Delhi has argued that existing laws were enough. The Information Technology Rules of 2021 already impose duties on platforms, MeitY issued AI advisories in 2024 that were quickly softened after industry pushback, and the India AI Governance Guidelines released in November 2025 were voluntary. In April the ministry proposed that platforms such as YouTube, Instagram and X continuously label AI-generated content. The new paper, Vaishnaw said, will be the basis for a dedicated AI law.
What the India AI regulation paper will cover
Speaking at the launch of the World Development Report, Vaishnaw said safety is “the most important point” and singled out deepfakes as a nuisance whose improving quality is misleading people and setting false narratives that can lead to harmful actions, according to Business Today. The consultation will also examine psychological harm to users, an area where he noted cases worldwide.
The ministry is reportedly exploring a consent-based framework for synthetic content, possible limits on how autonomously agentic AI systems may act, and regulatory sandboxes for high-risk applications. Vaishnaw described the approach as techno-legal, meaning technical controls paired with law rather than legislation alone, and said industry must carry the primary responsibility for cybersecurity, safety, user harm and deepfakes.
He also tied regulation to sovereignty, repeating a goal of domestically designed GPUs for AI inference within three to four years, and said no section of society should be left behind as AI-led change accelerates.
Why it matters
India has more than 900 million internet users and is one of the largest markets for ChatGPT, Gemini and Meta AI. A binding Indian AI law, especially one with limits on agentic autonomy and consent rules for synthetic media, would shape product decisions at every global lab, much as the EU AI Act did. It also lands as Singapore’s MAS finalizes its own AI risk rules and China expands its “AI Plus” program, making October a busy month for Asian AI policy.
The paper is due by early November. The details of what counts as a deepfake, and how hard the agentic limits bite, will decide whether this is light-touch or something closer to Brussels.