PIB TrackerIndia tightens the net around AI deepfakes
The government set out its framework against AI-generated deepfakes — IT Act offences, blocking and takedown powers, platform accountability and faster removal of unlawful synthetic content — alongside 13 responsible-AI projects for deepfake detection.
What happened
- The government set out its framework against AI-generated deepfakes — synthetic audio, video and text.
- The IT Act, 2000 covers identity theft (66C), impersonation (66D) and privacy violation (66E).
- Section 69A enables blocking orders; section 79 enables takedown notices to intermediaries.
- 13 responsible-AI projects were approved for deepfake detection under IndiaAI.
- Over 11.37 lakh people benefited from 6,650 cyber-security awareness workshops.
For Prelims
- Deepfake: AI-generated synthetic audio/video/text that convincingly impersonates real people.
- IT Act sections: 66C identity theft, 66D impersonation, 66E privacy, 67/67A obscene content.
- Section 69A: power to issue blocking orders to intermediaries.
- Section 79: safe harbour for intermediaries, conditional on due diligence and takedown on notice.
- IT Rules, 2021: impose due diligence, grievance officers and time-bound removal on platforms.
- IndiaAI: 13 responsible-AI projects for deepfake detection under the Safe & Trusted AI pillar.
For UPSC: A live tech-governance item. Use it for AI regulation and the IT Act/IT Rules, intermediary liability and safe harbour, threats to electoral integrity and privacy from synthetic media, and the detection-plus-awareness approach.
What it is NOT: This is enforcement under existing IT law plus detection projects and awareness — not a dedicated standalone deepfake or AI statute.
For Mains
Syllabus: GS3.18 · GS3.13 · Linkage L2
Anchor
When seeing is no longer believing — the state's answer to synthetic media.
Substantiation (data)
IT Act sections 66C/66D/66E/67, blocking under 69A and takedown under 79, plus 13 deepfake-detection projects and 6,650 awareness workshops reaching 11.37 lakh people.
Exemplification
Faster removal of unlawful AI content and strengthened grievance redressal put the onus on platforms as well as the state.
Problematisation
Detection lags generation; over-broad takedown risks free speech; and attribution across borders remains hard.
Way-forward
Combine provenance/watermarking standards, rapid but reviewable takedown, detection R&D, platform accountability and public media literacy.
Position
Deepfake governance must protect dignity, privacy and elections without becoming a licence for arbitrary censorship.
Deploys into: Cyber security + IT (GS3.18, GS3.13) · deepfakes and synthetic media, the IT Act/IT Rules and intermediary liability, and electoral and privacy risks from AI.
Ministry of Electronics & IT · 2026-07-30 · PRID 2291671 · PIB source ↗