India's Ministry of Electronics and Information Technology (MeitY) has proposed amendments to the Information Technology Rules that would require social media platforms and other intermediaries to remove unlawful AI-generated content within three hours of receiving a valid government or court order, replacing the current 36-hour deadline. The draft proposals also mandate the labelling of AI-generated content and faster grievance redress mechanisms as part of a broader effort to combat deepfakes, AI-driven misinformation and other harmful synthetic media.
According to MeitY, the proposed amendments are intended to address the growing threat posed by AI-generated deepfakes and other synthetic content, including manipulated audio, videos and text. The government said the stricter rules aim to curb the spread of unlawful AI-generated material while creating what it described as "an open, safe, trusted and accountable cyberspace for users."
KEY CHANGES PROPOSED UNDER THE DRAFT RULES
Three-hour deadline for removing unlawful AI content: Social media platforms would be required to take down unlawful AI-generated content within three hours of receiving a valid government or court notice. The current compliance window is 36 hours.
Mandatory labelling of AI-generated content: Platforms would have to clearly identify AI-generated images, videos, audio and other synthetic media. They would also be required to attach traceable metadata to help users distinguish AI-created content and minimise the risk of deception.
Stronger action against deepfakes: The draft rules specifically target deepfakes, AI-enabled impersonation, child sexual exploitation material, non-consensual intimate imagery and other harmful AI-generated content. Platforms such as Facebook, X and others would be required to act promptly against such material and deploy reasonable technical measures, including automated tools, to prevent its creation and circulation.
Faster grievance redressal: The government has proposed reducing the timeline for resolving general user complaints from 72 hours to 36 hours. Complaints involving sensitive issues, including nudity and impersonation, would need to be addressed within two hours instead of the existing 24-hour window.
Stricter responsibilities for platforms: The draft amendments would require social media companies to inform users about the legal consequences of sharing unlawful AI-generated content. Significant social media intermediaries would also have to implement appropriate technical measures to proactively detect and restrict the spread of certain categories of illegal content.
The proposed amendments are currently in the draft stage and have been released for public consultation. The government will consider feedback from stakeholders before finalising the revised rules, meaning the proposals may be modified before they are implemented.
