The model aims for regulatory compliance by design, embedding accountability into digital architectures and reducing manual enforcement.

Practical guidelines urge industry to comply with existing laws, publish transparency reports, implement grievance mechanisms, and favor a flexible, techno-legal oversight approach that balances innovation with harms mitigation.

India launches the India AI Governance Guidelines under the IndiaAI Mission to enable safe, inclusive, and responsible AI adoption across sectors, signaling a practical, people-first approach.

The framework rests on seven ethical principles built around a core “do no harm,” prioritizing human-centric development, risk mitigation, innovation sandboxes, and flexible governance that balances innovation with accountability.

MeitY emphasizes leveraging AI for public benefit while keeping regulation as a fallback, ready to act if necessary but not the priority at this stage.

Data protection remains central, with debates over training on public data, purpose limitation, consent managers, and legitimate use exemptions, underpinned by calls for legislative review.

The policy pillar advocates agile regulation, prioritizing existing laws and targeted amendments over an immediate AI-specific act, with clarity sought on intermediary liability and how AI systems are classified.

User consent and data transparency are stressed, noting that data used to train AI models should comply with the Digital Personal Data Protection Act.

The guidelines are non-binding but aimed at shaping future regulation, signaling a cautious, principle-based approach rather than rushing new laws.

Deepfakes, content provenance, and copyright implications are addressed, with consideration of a national content verification commission.

Safeguards against misuse of foundation models and deepfakes are acknowledged, with emphasis on traceability through unique identities for creators, publishers, and platforms to enable watermarking and lifecycle tracking.

Recommendations include agile, principle-based reform, targeted amendments to copyright and data protection, standards for authentication and fairness, global coordination on authentication, regulatory sandboxes with immunities, and horizon scanning.