Analysis
Tick, tock, takedown
Do the new three-hour digital takedown rules undermine the limits Supreme Court decisions have placed on executive power?

Earlier this week, responding to deepfake proliferation on social media, the Ministry of Electronics and Information Technology (MeitY) notified amended Information Technology Rules, effective from 20 February, with some changes from the draft version released in October last year.
The key change is the compression of the timeline within which internet intermediaries have to comply with the government’s takedown orders: from 36 hours to three hours, and from 24 hours to two hours for non-consensual intimate imagery. The time for grievance redressal is now seven days—down from 15. Timelines for resolving and removing information upon “actual knowledge” or on the basis of a complaint have also been shortened.
The government has framed the amendments as necessary to combat the perils of synthetic media. In an article published by the Hindustan Times on Wednesday, MeitY Secretary S. Krishnan described the changes as part of India’s wider push for responsible AI, a move from “reactive moderation to ex ante governance.” Anticipating the pushback from Big Tech, Krishnan wrote that the regime is “not an assertion of regulatory excess, but a reaffirmation of democratic stewardship.”
Yet, concerns that the rules could have a chilling effect on free speech continue to swirl. Previously, the Supreme Court has been called in to draw the line between intermediary liability and free expression. In Shreya Singhal v Union of India (2015), the Court struck down Section 66A of the IT Act for its chilling effect on speech, holding that vague standards and sweeping takedown powers violated Article 19(1)(a). In Puttaswamy v Union of India (2017), the Court held that restrictions on fundamental rights must be by the least intrusive means available. In K.A. Abbas v Union of India (1970), the Court had noted that the reasonableness of a restriction must be assessed with reference to its nature and extent, not merely its stated purpose.
The Supreme Court has repeatedly affirmed that the right to free speech includes the right to disseminate information. In Indian Express Newspapers v Union of India (1984), the Court held that restrictions on circulation undermine the right at its source.
And, in Anuradha Bhasin v Union of India (2020), a case relating to web blackouts in Jammu and Kashmir, the Court held that internet restrictions must be temporary, proportionate and subject to review.
Arguably, the three-hour compliance window inverts the Shreya Singhal framework, which requires specificity and judicial oversight before content removal. The compressed timelines favour precautionary takedowns. Where content involves political speech, satire or regional language requiring contextual interpretation, platforms now face a difficult choice: remove first and review later, or risk losing safe harbour protection under Section 79 of the IT Act.
Legal experts have noted that when intermediaries must act within hours, automation and rule-based enforcement displace human judgement. Edge cases—parody involving public figures, remixed news content, commentary during elections—could become too risky to host. A three-hour window forecloses a proper internal review, legal consultation or consideration of defences, besides raising issues around compliance-feasibility, particularly for smaller firms.
Industry observers have also warned that the amendments incentivise platforms to adopt conservative moderation policies favouring removal over assessment. As we speak, lawyers at intermediary companies may be considering whether the new compliance window satisfies the Puttuswamy threshold of ‘least intrusive method’, particularly where content may qualify for protection under fair use, satire or political commentary. The unequivocal regulatory signal, however, is that speed of enforcement takes precedence over depth of assessment.
The timing of the amendment is noteworthy. It arrives as the Government of India positions itself as a proponent of innovation-friendly AI governance in the run-up to the splashy AI Impact Summit in Delhi next week. While the truncated consultation and lack of parliamentary debate on the changes will likely be whispered about behind closed doors in the capital, it remains to be seen whether the government will be legally challenged on the rules and their hasty passage.
This article was first featured in SCO’s Weekly newsletter. Sign up now!