Policing Deepfakes: Inside India's First AI Rulebook

On December 4, 2025, the Supreme Court quietly stepped away from a sweeping petition asking it to create a licensing regime for AI systems and deepfake tools.

- AGENCIES

Representative image.

The reason: the Union government told the bench it had already drafted an AI rulebook and put it out for public comment.

In Aarati Sah v. Union of India, petitioners had sought a national authority to license powerful AI models, mandatory labelling of synthetic content, and rapid takedowns of non-consensual deepfakes. The Court, led by Chief Justice Surya Kant, noted that draft rules were now in the public domain and disposed of the PIL, signalling that the regulatory battle would move from courtroom to rule-making table.

The draft rules: deepfakes in plain English

Parallel to the PIL, the IT Ministry has proposed amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 specifically targeting “synthetically generated information.”

@ A new legal definition: “Synthetically generated information” is any content artificially created or altered by computer to appear reasonably authentic—essentially, deepfakes and Al-generated media.