India’s draft IT rules

India just put out a draft that could pull a fake video, a political meme, and even a user-written correction on X into the same regulatory system that already governs digital news publishers. The Ministry of Electronics and Information Technology published the new draft on March 30, 2026, and asked for comments by April 14, 2026. (meity.gov.in) One part of this story is old and one part is new. India had already amended its Information Technology Rules on February 10, 2026 to require labels and provenance markers for “synthetically generated information,” and the March 30 draft goes further by widening platform duties and government oversight. (meity.gov.in, meity.gov.in) The February rules defined synthetic content broadly as material artificially created, generated, modified, or altered by a computer in a way that appears reasonably authentic. The explanatory note lists the targets as deepfakes, fabricated political clips, financial impersonation, and other media that can look real enough to fool people. (meity.gov.in) That earlier change is why platforms in India are already dealing with AI labels. The government’s own frequently asked questions page says the amended 2026 rules were notified on February 10, 2026, after an October 22, 2025 draft focused on synthetic media. (meity.gov.in, meity.gov.in) The March 30 draft changes the power balance in a different way. It would insert a new Rule 3(4) that treats ministry clarifications, advisories, directions, standard operating procedures, codes of practice, and guidelines as part of the due-diligence duties platforms must follow to keep their legal protection. (meity.gov.in) That legal protection is the shield that normally keeps a platform from being treated as the speaker of every user post. Open and The Hindu both report that missing the new deadlines or duties could put “safe harbor” at risk for companies such as X, Meta, and YouTube. (openthemagazine.com, thehindu.com) The deadlines are much shorter than the older internet norm of “we’ll review it.” The Hindu reported in February that platforms must act on government or court orders within three hours, down from 36 hours, and Open reported that some impersonation cases could shrink to two hours. (thehindu.com, openthemagazine.com) The other big shift is who counts as regulated media. The March 30 notice says Part III of the rules would be clarified to apply to intermediaries and to news and current-affairs content hosted by non-publisher users, which means ordinary users can be pulled into a framework once aimed at formal publishers. (meity.gov.in) That is where X’s Community Notes feature enters the picture. Hindustan Times reported on April 10 that the government wants Community Notes touching news, politics, or public policy to fall under the Ministry of Information and Broadcasting’s remit, so a user-written note correcting an official claim could face removal through the same system. (hindustantimes.com) The report tied that push to actual posts, not a hypothetical fight. Hindustan Times said Community Notes had appeared this year on posts by Prime Minister Narendra Modi, education minister Dharmendra Pradhan, and information technology minister Ashwini Vaishnaw, and some of those notes were later removed. (hindustantimes.com) Put together, the rules create two separate obligations at once. Platforms would have to label and rapidly remove synthetic media, and they would also have to treat a wider range of user speech, including some moderation features built by users themselves, as content that ministries can scrutinize or escalate. (meity.gov.in, meity.gov.in, hindustantimes.com) The draft is not final yet, but the direction is already clear in the dates. India first tightened AI and deepfake rules on February 10, 2026, then moved on March 30, 2026 to widen oversight beyond fake media itself and into the systems platforms use to host, rank, label, and correct speech. (meity.gov.in, meity.gov.in)