India has mandated social media platforms to enhance their monitoring of deepfake content and other AI-generated impersonations, significantly reducing the timeframe for compliance with takedown requests. This initiative is poised to transform content moderation practices for global tech companies within one of the fastest-growing internet markets.
The newly revised regulations, introduced as amendments to India's 2021 IT Rules, officially categorize deepfakes under a structured regulatory framework. This includes requirements for the labeling and traceability of synthetic audio and visual materials, with strict deadlines imposed on platforms--three hours for official takedown requests and two hours for urgent user complaints.
Given India's substantial digital user base, exceeding one billion, and its youthful demographic, these rules are expected to significantly influence how major platforms like Meta and YouTube operate, potentially affecting global moderation standards.
As per the amended guidelines, social media platforms permitting user-uploaded audio-visual content must ensure clear disclosures regarding the synthetic nature of such materials, utilize verification tools, and label deepfakes with traceable provenance data.
Certain types of synthetic content, such as misleading impersonations and non-consensual intimate imagery, are strictly prohibited. Non-compliance, especially in cases highlighted by authorities or users, may increase legal liabilities for companies, risking their safe-harbor protections under Indian law.
The regulations place a strong emphasis on automated systems to fulfill these responsibilities. Platforms are expected to implement technical solutions to verify user disclosures, identify and label deepfakes, and prevent the creation or dissemination of banned synthetic content.
Rohit Kumar, a founding partner at The Quantum Hub in New Delhi, remarked, "The revised IT Rules represent a more nuanced strategy for managing AI-generated deepfakes. The drastically shortened grievance timelines will significantly heighten compliance challenges and require careful examination, particularly since non-compliance could result in the loss of safe harbor protections."
Aprajita Rana, a partner at AZB & Partners, noted that the focus has shifted to AI-generated audio-visual content rather than all online information, allowing exceptions for standard, cosmetic, or efficiency-related AI applications. However, she expressed concerns that the three-hour removal requirement upon awareness diverges from traditional free-speech principles.
Critics, including the Internet Freedom Foundation, have voiced apprehensions that these regulations may expedite censorship by compressing takedown timelines, thereby limiting opportunities for human review and pushing platforms towards excessive automated removals. They also raised issues regarding the expansion of prohibited content categories and provisions allowing platforms to disclose user identities to complainants without judicial oversight.
Industry insiders indicated that the amendments followed a limited consultation process, with only a few suggestions incorporated into the final rules. While the Indian government has narrowed the scope to focus on AI-generated content, other recommendations were overlooked, suggesting a need for further consultation to clarify compliance expectations.
The revised rules are set to take effect on February 20, providing platforms with limited time to adapt their compliance systems. This rollout coincides with India hosting the AI Impact Summit in New Delhi, where global technology leaders and policymakers are expected to gather.
