MeitY pushes AI content labelling and watermarking to curb deepfakes and cybercrime, says IndiaAI Mission CEO
India’s IT ministry is working toward requirements for watermarking and labelling AI-generated content, citing misuse ranging from cybercrime to deepfakes and abusive imagery. The move would place a compliance burden on platforms and aims to make synthetic media easier to identify in everyday online sharing.
The Ministry of Electronics and Information Technology (MeitY) is working on mandating watermarking and labelling for AI-generated content, according to remarks attributed to Abhishek Singh, additional secretary at MeitY and CEO of the IndiaAI Mission. The effort is positioned as a response to the growing misuse of generative AI for cybercrime and deepfakes.
In the reported comments, Singh said the government had issued draft provisions related to watermarking and labelling, arguing that AI-generated content is increasingly being used to power fraud, create deepfakes, and produce abusive material including sexually explicit imagery. The stated goal is to make it clearer to users whether what they are viewing is synthetic.
The policy direction reflects a broader trend: shifting from voluntary platform policies toward legal duties that require clearer identification, traceability and accountability for synthetic media. In practical terms, watermarking and labelling can be implemented as visible indicators, metadata, or both—though enforcement details depend on final rules and compliance standards.
For platforms, the likely impact is operational: building or buying detection and provenance tooling, updating upload flows, and creating user-facing notices and reporting mechanisms. For users, the promise is a simpler signal—especially during elections, crises, and viral moments—about whether content may have been algorithmically created or altered.
MeitY’s draft provisions are part of India’s evolving approach to regulating online harms, with deepfakes and synthetic content now treated as a mainstream governance issue rather than a niche technical problem. Stakeholder feedback and implementation feasibility will shape how the final mandate is enforced across apps, social platforms, and creation tools.