India has stepped up its crackdown on AI-generated misinformation by ordering social media platforms to act far more quickly against deepfakes, a move that could redefine how content is moderated in one of the world’s largest digital markets.

The amended IT Rules, published on Tuesday, bring India deepfake rules into sharper focus by drastically reducing takedown timelines. Platforms must now comply with official government orders within three hours and resolve certain urgent user complaints within just two hours.

Deepfakes Brought Under Formal Regulation

The updated rules amend India’s 2021 IT framework and formally place synthetic audio and visual content under regulatory oversight. Platforms hosting user-generated audio-visual material must require disclosures on whether content is AI-generated, verify those claims, and ensure deepfakes are clearly labelled with traceable provenance data.

These obligations are designed to make synthetic content easier to identify while discouraging misuse at scale.

Why India’s Market Makes These Rules Global

India’s scale significantly amplifies the impact of the new framework. With more than a billion internet users and a predominantly young online population, the country is a critical market for global platforms.

As a result, compliance systems built to meet India deepfake rules are likely to influence product design and moderation practices well beyond the country’s borders.

Outright Bans and Higher Legal Risk

The rules prohibit certain categories of synthetic content altogether, including deceptive impersonations, non-consensual intimate imagery, and material linked to serious crimes.

Non-compliance, particularly in cases flagged by authorities or users, can jeopardise platforms’ safe-harbour protections under Indian law, exposing companies to greater legal liability.

Automation Takes Centre Stage

To meet the compressed timelines, platforms are expected to rely heavily on automated systems to detect, label, and prevent the creation or sharing of prohibited synthetic content.

Legal experts say this reliance on automation is unavoidable given the speed required under the revised India deepfake rules, but it also raises questions about accuracy and oversight.

Legal Experts Flag Compliance Pressures

Rohit Kumar, founding partner at New Delhi-based policy consultancy The Quantum Hub, described the amendments as a calibrated approach to regulating AI-generated deepfakes. However, he warned that the two- to three-hour takedown windows would materially raise compliance burdens, especially since non-compliance is tied to the loss of safe-harbour protections.

Aprajita Rana, a partner at AZB & Partners, noted that the rules now focus specifically on AI-generated audio-visual content rather than all online information, while carving out exceptions for routine or efficiency-related uses of AI. She cautioned, however, that the three-hour removal requirement departs from established free-speech principles.

Free Speech and Over-Removal Concerns

Digital rights group Internet Freedom Foundation warned that the India deepfake rules risk accelerating censorship by leaving little room for meaningful human review. The group argued that such short timelines could push platforms toward automated over-removal.

It also raised concerns about expanded prohibited content categories and provisions that may allow disclosure of user identities to private complainants without judicial oversight.

Limited Consultation, Tight Deadline

Industry sources said the amendments followed a limited consultation process, with only a narrow set of recommendations reflected in the final rules. While the government narrowed the scope to AI-generated audio-visual content, other proposals were not adopted despite significant changes between the draft and final versions.

The amended rules come into effect on February 20, giving platforms little time to adapt compliance systems. The rollout coincides with India hosting the AI Impact Summit in New Delhi from February 16 to 20, placing the country’s evolving deepfake framework firmly under global scrutiny.