India’s deepfake challenge calls for smart regulation, not heavy-handed rules
The opinion piece argues that deepfakes are spreading quickly and causing real harm in India, from scams and impersonation to broader trust erosion, but warns that the government’s proposed regulatory response risks going too far. It cites examples of convincing synthetic videos and references survey findings showing wide exposure to deepfakes and meaningful rates of victimization through deepfake enabled scams.
It criticizes proposed amendments to India’s Intermediary Guidelines that would create a category of “synthetically generated information” and impose strict labeling and verification duties on platforms. The article says the rules would require prominent visible labels and embedded metadata, and would force large platforms to collect user declarations about whether uploads are synthetic and then attempt to verify those claims with technical tools, with platforms labeling content themselves if users do not.
The author’s core concern is that this would undermine India’s safe harbour model under Section 79 of the IT Act by pushing intermediaries to inspect, classify, and modify user content before it is published, effectively treating them more like publishers. The piece argues this shift is both legally fraught and practically unworkable because deepfake detection is still unreliable, and compliance pressure would lead platforms to block or delay legitimate speech to reduce liability, especially harming smaller services that cannot absorb the cost.
Instead, it calls for a more targeted approach focused on provenance and user resilience. It points to provenance systems and creation time watermarking as more realistic tools than platform inserted labels, argues that public awareness and digital literacy should be central, and recommends preserving safe harbour while using fast track remedies and coordinated enforcement for clear criminal misuse like fraud, impersonation, and reputational attacks.
