Guidelines, advisories, circulars, standards and court AI policies. Most are not legally binding; each entry states its legal effect.
The Indian Patent Office's rules for how its own examiners and Controllers may, and may not, use AI tools, including a ban on entering unpublished applications into public AI
Status checked Oct 15, 2026
The drug regulator's final guidance on licensing software as a medical device, with explicit requirements for AI and machine learning tools
Status checked Oct 15, 2026
Gujarat High Court bars AI from adjudication and drafting of orders, fixes personal liability on signatories, and later enforces AI norms on State tax officers
Status checked Oct 15, 2026
MeitY's notice to X over Grok-generated obscene images, with a 72-hour action-taken report and a warning of lost safe harbour
Status checked Oct 15, 2026
MeitY's framework of seven 'sutras', six pillars and new institutions that governs AI through existing laws rather than a standalone AI statute
Status checked Oct 15, 2026
BIS lists 26 published AI and big-data Indian Standards adopted from ISO/IEC, including the AI management system standard IS/ISO/IEC 42001:2023
Status checked Oct 15, 2026
Reported as India's first High Court AI policy for judges: no AI for findings or judgments, no cloud chatbots, audit logs and mandatory training
Status checked Oct 15, 2026
The Patent Office's 2025 examination guidelines for software and AI inventions, with a dedicated chapter on AI/ML, inventorship and disclosure
Status checked Oct 15, 2026
ECI requires parties to label AI-generated campaign material, with labels covering 10% of the screen, creator disclosure, record-keeping and three-hour takedowns
Status checked Oct 15, 2026
QCI's accreditation board sets the criteria under which bodies may certify organisations against ISO/IEC 42001, the AI management system standard
Status checked Oct 15, 2026
The advisory that first required government permission for 'under-tested' AI models, drew a start-up backlash and was rewritten two weeks later
Status checked Oct 15, 2026
MeitY's first deepfake advisories, issued after the Rashmika Mandanna video, warned platforms that failing to act would cost them safe harbour
Status checked Oct 15, 2026
The Telecommunication Engineering Centre's voluntary standard for testing AI systems for bias and issuing a fairness score
Status checked Oct 15, 2026
A sector-specific ethical framework for AI in health research and care, issued by the Indian Council of Medical Research
Status checked Oct 15, 2026
NITI Aayog's two-part approach document setting seven principles for responsible AI in India, grounded in constitutional morality
Status checked Oct 15, 2026