India orders social media platforms to take down deepfakes faster

February 10, 2026

TL;DR

India has updated its IT Rules to address deepfakes and other AI-generated impersonations.
Social media platforms must label synthetic audio and visual content and embed traceable provenance data.
Takedown deadlines for platforms have been drastically reduced, with a three-hour window for official orders and a two-hour window for urgent user complaints.
Certain categories of synthetic content, such as deceptive impersonations and non-consensual intimate imagery, are strictly prohibited.
Non-compliance risks jeopardizing companies' safe-harbour protections under Indian law.
The rules emphasize the use of automated systems for verification, identification, labelling, and prevention of prohibited synthetic content.
Critics argue the compressed timelines could accelerate censorship, reduce human review, and push platforms toward automated over-removal, potentially undermining free speech and due process.
The amendments followed a limited consultation process, and concerns have been raised about the lack of judicial oversight for user identity disclosure to private complainants.
The new rules come into effect on February 20, shortly before India hosts the AI Impact Summit.

Continue reading
the original article