India’s government has moved to tighten rules governing AI-generated material on social media, imposing a dramatically shortened timeline for platforms to remove content deemed illegal or harmful and mandating permanent labels for synthetic media. According to the announcement, the changes take effect on 20 February 2026, the final day of an international AI summit in New Delhi, and cut the window for complying with government takedown notices from 36 hours to three. (Sources: India Today, New Age).

The measures apply to major global services including Instagram, Facebook (Meta), YouTube and X and broaden the definition of regulated content to include material “created, generated, modified or altered through any computer resource”, excluding routine or “goodfaith” editing. Industry observers and legal analysts say the amendments mark the first formal regulation of AI-manipulated content under India’s intermediary rules. (Sources: Times of India, New Age).

The government is also requiring platforms to obtain declarations from users when content is AI-assisted, to label synthetic media with markings that cannot be removed or suppressed, and to deploy automated tools to detect and block illegal material such as forged documents, child sexual abuse imagery and other criminal content. Government filings describe these measures as necessary to curb the rapid spread of disinformation and sexualised imagery facilitated by increasingly accessible AI tools. (Sources: India Today, Law analysis).

Digital rights groups have warned the compressed notice period will force platforms into what they call hasty removals and could concentrate control away from users. Apar Gupta of the Internet Freedom Foundation warned the timelines are “so tight that meaningful human review becomes structurally impossible at scale” and argued the system shifts decision-making “decisively away from users”, with grievance and appeals processes operating on slower clocks. (Sources: New Age, India Today).

Critics argue the rules risk sweeping in legitimate speech, including satire, parody and political commentary that use realistic synthetic media. “It is automated censorship,” digital rights activist Nikhil Pahwa told AFP, and the US-based Center for the Study of Organized Hate, in a report with the Internet Freedom Foundation, cautioned that proactive monitoring could produce collateral censorship as platforms err on the side of removal. Observers further note that labelling and metadata-based approaches are technically fragile because metadata can be stripped when content is edited, compressed, screen-recorded or cross-posted. (Sources: New Age, CSOH/IFF report).

Supporters of the rules say tighter enforcement was compelled by repeated episodes in which synthetic tools were used to produce harmful imagery and disinformation at scale, citing recent controversies where generative systems enabled mass production of sexualised images and manipulated media. Government and some civil-society voices frame the amendments as an attempt to make platforms more accountable for preventing demonstrable harms online. (Sources: India Today, TechCrunch).

Implementation will test the balance between rapid removal of dangerous content and protection of free expression in the world’s largest democracy. Legal experts say the practical challenges of verifying vast volumes of synthetic material, the technical limits of reliable detection, and the broad wording of takedown criteria leave substantial room for differing interpretations and potential legal challenge as the rules come into force. (Sources: Business Today, Roya, Times of India).

Source Reference Map

Inspired by headline at: [1]

Sources by paragraph:

Source: Noah Wire Services