IT Rules Amendment & AI Regulation (2026)

IT Rules Amendment & AI Regulation (2026)

Context

The Ministry of Electronics and Information Technology (MeitY) officially notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026. These rules, set to come into force on February 20, 2026, introduce the first formal regulatory framework for "Synthetically Generated Information" (SGI) to combat the surge of deepfakes and AI-led misinformation.

Key Amendments (2026)

• Mandatory AI Labeling: * Platforms must ensure that all synthetic content (AI-generated or altered) carries a clear and prominent label.
  • Traceability: Intermediaries must embed permanent metadata or unique identifiers (provenance markers) to trace the content back to the source platform.
  • Note: A previous proposal to mandate that labels cover exactly 10% of the screen was dropped in the final notification following industry feedback.
• Drastic Takedown Windows:
  • Lawful Orders: Platforms must remove content flagged by a court or government order within 3 hours (down from 36 hours).
  • Sensitive Content: Non-consensual deepfakes or intimate imagery must be removed within 2 hours (down from 24 hours).
• User Declarations: Significant Social Media Intermediaries (SSMIs) must now require users to declare if their content is AI-generated at the time of upload. Platforms are also expected to deploy technical tools to verify these declarations.

Key Concepts & Definitions

• Synthetically Generated Information (SGI): Defined as any audio, visual, or audio-visual information artificially created or modified to appear real/authentic.
  • Exemptions: Routine editing (color correction, noise reduction), accessibility tweaks (transcription/translation), and good-faith academic or training materials are not classified as SGI.
• Safe Harbour (Section 79): The 2026 amendment clarifies that platforms will retain their legal immunity if they remove synthetic content in good faith. However, they lose this protection if they "knowingly permit or promote" unlabelled SGI or fail to act within the new 2-3 hour window.

Comparison of Takedown Timelines

Concerns & Challenges

• The "Chilling Effect": Critics argue that the 3-hour window is too short for platforms to distinguish between malicious deepfakes and legitimate political satire or criticism, potentially leading to over-censorship.
• Verification Feasibility: While platforms must verify user declarations, technical experts warn that detecting high-quality "stealth" deepfakes in real-time remains a significant engineering challenge.
• Algorithmic Bias: Automated moderation tools used to meet these tight deadlines may inadvertently flag regional dialects or cultural nuances as "synthetic."

Conclusion

The 2026 IT Rules represent a shift toward "Safety-by-Design." By mandating provenance and near-instant removal, India is moving away from purely reactive moderation to a system of active accountability for both AI tool providers and social media giants.