India Mandates 3-Hour Takedown for Deepfakes
India has implemented a new amendment to its IT rules requiring platforms to remove illegal deepfakes and other specified AI-generated content within three hours of a government or court order. The rule shortens to two hours for content involving sexual acts or impersonation. All AI-generated media must now also feature tamper-proof metadata and visible labels to indicate its origin.
- The rules, which took effect on February 20, 2026, amend the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. They were notified by the Ministry of Electronics and Information Technology (MeitY) on February 10, giving platforms ten days to comply. - This amendment introduces a legal definition for "synthetically generated information," covering any AI-altered media that appears authentic. The regulation was prompted by a surge in viral deepfake videos targeting politicians and celebrities, raising concerns about misinformation and reputational damage. - Non-compliance can lead to the loss of "safe harbor" protection under Section 79 of the IT Act, exposing platforms to civil and criminal liability as publishers of the content. Penalties can include fines or, in extreme cases, the blocking of non-compliant platforms. - Platforms must now warn users every three months, instead of annually, about the consequences of creating or sharing unlawful content, which can include penalties under the Indian Penal Code and the IT Act. - Compared to other major regulatory frameworks, India's timeline is significantly more aggressive. The EU's AI Act, with mandatory compliance expected from August 2026, requires labeling but doesn't specify takedown times, while the U.S. TAKE IT DOWN Act has a 48-hour window for non-consensual intimate deepfakes. - The initial draft of the rules proposed a watermark covering at least 10% of AI-generated visuals, but the final version adopted a more flexible "prominent and visible" labeling standard after industry feedback. - Platforms are now required to use automated tools to verify user declarations about whether their content is AI-generated, moving beyond simple self-declaration. - Concerns have been raised by digital rights advocates and some industry experts that the extremely short takedown window could lead to over-censorship of legitimate content, including satire and parody, as platforms act hastily to avoid penalties.
