The Ministry of Electronics and Information Technology scheduled a week of labeling clinics for platforms and creator economy teams to explain synthetically generated information duties that took force in February, using Cyber Security Awareness Month in October to push compliance beyond legal desks into product and trust-and-safety squads.
What the 2026 rules require
Amendments notified on 10 February 2026 under the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, require intermediaries to prevent unlawful AI-generated content, including non-consensual intimate imagery and deceptive impersonation, with technical safeguards, labelling, and prompt removal pathways. MeitY’s FAQ stresses that platforms must not wait for court orders when harm is evident.
The ministry distinguishes between risk-based India AI Governance Guidelines released in November 2025 and binding due diligence on intermediaries. Startups building generative tools must still map whether they qualify as intermediaries or significant social media intermediaries, a classification that changes reporting timelines.
October outreach
Officials plan hybrid sessions in Delhi, Bengaluru, and Hyderabad covering metadata labels, visible disclosures, and grievance officer workflows. Cyber Jagrookta Diwas events on the first Wednesday of each month will include demos of deepfake detection pilots funded under the IndiaAI Mission’s Safe and Trusted AI pillar, according to a Press Information Bureau note summarising government responses in Parliament.
Industry pressure
Entertainment unions and actors pushed for faster takedowns after AI-altered clips of performers circulated online this week. MeitY officials told clinic registrants that amended rules already cover sexually explicit synthetic media, but enforcement depends on platforms implementing upload filters and user reporting queues that scale in Indian languages.
Technical expectations
Workshops will review watermarking standards, provenance metadata, and model-card disclosures for government procurement. Agencies issuing chatbot tenders after October must log training-data sources—a requirement Taiwan’s desk also adopted, but India’s version ties to MeitY-administered rules rather than sector regulators alone.
Research pipeline
Thirteen responsible-AI projects at universities include deepfake detection, per PIB. MeitY said it will pair those tools with platform APIs where memoranda of understanding exist, though no mandatory API access is spelled out in the FAQ.
Global context
Other jurisdictions moved on synthetic media in 2026, but India’s intermediary liability frame means compliance work lands on large platforms first, with over-the-top media and messaging apps facing the strictest timelines. Smaller SaaS vendors are advised to document whether they host user-generated content or merely provide enterprise APIs.
What attendees should bring
Product counsel are urged to carry current moderation playbooks and Hindi, Tamil, and Bengali support metrics. MeitY teams said clinics will not certify compliance but will flag common gaps—missing SGI labels on political speech during state campaigns, or delayed responses on intimate-image abuse reports.
Next checkpoints
Awareness month ends 31 October, but rule obligations are continuous. Platforms that treat clinics as one-off PR risk audit letters when CERT-In or MeitY review grievance logs. For AI teams, the practical output is a labelled-content pipeline ready before Diwali viral marketing peaks.