AI Consultation Paper Due Within a Month; What the Deepfake Rules Already Require
Ashwini Vaishnaw said on 8 October 2026 the paper will cover AI safety and deepfakes. The February IT Rules already set three-hour takedowns and synthetic media labels.
Union IT Minister Ashwini Vaishnaw said on 8 October 2026 that the government will publish a consultation paper on regulating AI within a month. He made the remark at an event in New Delhi marking the release of the World Development Report. The one-month window points to about 8 November 2026, but no fixed date has been given.
The paper comes after a set of rules India already has for AI-generated content. In February 2026 the Ministry of Electronics and Information Technology (MeitY) amended the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. The changes require labels on synthetic media, including deepfakes, and cut the deadline for removing unlawful content to three hours. The paper is set to form the foundation for later AI regulation. For platforms, startups and users, the distinction matters: the February duties are in force now, while the paper's contents are not yet public.
What the Consultation Paper Is Expected to Cover
Vaishnaw named AI safety, deepfakes, harms linked to AI use, a human-first approach and skilling as focus areas. BusinessToday also says the consultation will consider psychological impacts on users, which ThePrint's list of focus areas does not include. No detail on the human-first approach has been announced.
On deepfakes, he called them a nuisance that can cause significant problems in society. He said their improving quality misleads people and can set wrong precedents and narratives that may lead to harmful actions.
He said the industry has to take primary responsibility for issues such as cybersecurity, safety, user harm, psychological disorders and deepfakes. “We have to bring a techno-legal solution,” Vaishnaw said. No list of the technical measures involved has been announced.
On skilling, he stressed that no section of society should be left behind as AI-led change speeds up. Whether the paper will be a draft for public comment has not been announced.
The February 2026 Amendment and Its Effective Date
MeitY notified the amendment on 10 February 2026, and it took effect on 20 February, giving intermediaries a 10-day compliance window.
The amendment changes Rule 3(1)(d) and adds labelling duties under Rule 3(3) of the existing IT Rules. It defines synthetically generated information as audio, visual or audio-visual content that a computer resource artificially or algorithmically creates or alters in a manner that appears real, authentic or true. The definition also covers content that depicts or portrays a person or event in a way that is, or is likely to be perceived as, indistinguishable from a real one.
The amendment followed an October 2025 draft. Medianama reports that the final text is narrower in scope than that draft, which applied to information without limiting its form.
The consultation paper's listed focus returns to deepfakes, the subject the February amendment already addressed.
A Three-Hour Deadline for Removing Unlawful Content
According to Medianama, the amended Rule 3(1)(d) cuts the removal window for unlawful content from 36 hours to three. Hogan Lovells says intermediaries must act on government or court orders, including takedown orders, within three hours of receipt.
The window is three hours. Medianama also reports that the final rules clarify safe harbour protections for removals carried out in compliance with them.
Hogan Lovells presents the three-hour deadline as covering illegal content, while its labelling requirements cover synthetic media. It also lists the categories of synthetic content the rules treat as prohibited: child sexual exploitation material, non-consensual intimate imagery, obscene or sexually explicit content, false documents or electronic records, synthetic depictions of explosives or arms, and deceptive portrayals of individuals or events.
Labels, Provenance Markers and the Scope of Synthetic Content
Intermediaries must prominently label synthetic content in a way that is easily noticeable, and embed permanent metadata or other appropriate technical provenance mechanisms, to the extent technically feasible, Medianama reports. Hogan Lovells says platforms cannot enable modification, suppression or removal of these markers.
For audio, the rules require a prominently prefixed disclosure. A unique identifier must also identify the intermediary's computer resource used to create or alter the content.
Medianama reports that the final rules dropped the mandatory 10% label-size requirement from the October 2025 draft in favour of a prominence standard.
On scope, Medianama and Hogan Lovells both describe the definition as covering audio, visual and audio-visual content, and both treat it as turning on content that appears real and authentic.
“We have to bring a techno-legal solution,” Vaishnaw said.
What Platforms and Startups Need to Prepare For
The labelling duty is placed on intermediaries under Rule 3(3). Platforms above the significant social media intermediary threshold carry additional duties. Hogan Lovells puts that threshold at more than 5 million registered users in India.
Hogan Lovells says in-scope platforms must implement detection tools and labelling systems, and should strengthen monitoring, ahead of the 20 February 2026 effective date. That date has passed, so these obligations are already in effect.
For a company offering tools that generate or alter images, audio or video, Hogan Lovells says such tools must deploy technical measures against prohibited synthetic content. For a platform that hosts user uploads, it says escalation protocols should be reviewed for the three-hour deadline.
The two sets of duties differ in timing: the labelling and takedown rules took effect in February, while no publication date has been set for the paper.
Vaishnaw's remarks placed primary responsibility for deepfakes and user harm on industry, and deepfakes are among the paper's focus areas.
Timeline and What Has Not Been Announced
The one-month window from 8 October runs to about 8 November 2026. No fixed publication date has been announced.
It has not been announced whether the paper will propose changes to the IT Rules, new legislation, or both, though BusinessToday reports the government is now moving towards a dedicated AI law. The listed focus includes skilling.
Until the paper is published, the February amendment is the operative text on synthetic media, and its labelling, metadata and takedown duties are the ones in force. The one-month timeline is the government's stated plan, with no set date.
