On November 5, 2025, India’s Ministry of Electronics and Information Technology (MeitY) launched the India AI Governance Guidelines (IAGG) as part of the IndiaAI Mission. This is a big step to make AI ethical and fair for everyone in India.
A team led by MeitY created these guidelines using ideas from global rules, India’s laws, and public feedback. It uses a practical mix of technology and law. The focus is on voluntary steps, rewards, and shared digital tools (DPI) to encourage innovation while reducing risks. This fits India’s “AI for All” goal and dream of becoming a developed country (Viksit Bharat) by 2047.
Main Goals :-
IAGG wants to spread AI benefits to areas like health, education, farming, and government. It will help grow the economy, make society stronger, and compete globally.
It tackles problems like AI biases, fake videos (deepfakes), data leaks, and security threats. To build trust, it stresses human-focused design that fits India’s many languages and cultures. Using IndiaAI Mission tools—like 38,231 GPUs and over 1,500 datasets on AIKosh—it builds strong, home-grown AI without too many strict rules.
Seven Key Principles (Sutras)
Based on RBI’s FREE-AI ideas, these simple rules guide everything (they work for any tech):
Trust First: Make AI reliable from start to finish.
People Come First: Keep humans in control and empowered.
Innovate Smartly: Get the most good from AI while cutting down harms.
Fairness for All: Include everyone, especially weaker groups.
Who’s Responsible?: Clearly say who handles risks.
Easy to Understand: Make AI clear and explainable.
Safe and Green: Build strong, secure, and eco-friendly systems.
Main Parts and Suggestions
The guidelines cover six areas for support, rules, and checks:
Building Tools and Skills: Give more data and computing power. Train over 500 PhDs through IndiaAI FutureSkills. Run campaigns to teach regulators and people about AI.
Rules and Policies: Use current laws (like IT Act 2000 and DPDP Act 2023) for AI issues like deepfakes. Suggest changes for easier data use and watermarks on content.
Reducing Risks: Set up a national database for AI mistakes. Require human checks in key steps. Reward companies for self-checks.
Responsibility and Groups: Start AI Governance Group (AIGG) for teamwork, Technology & Policy Expert Committee (TPEC) for future planning, and AI Safety Institute (AISI) for tests and standards. Groups like RBI and TRAI will enforce rules in their areas.
There’s a plan with quick wins (like starting AIGG in 9-12 months), medium steps (like test zones), and long-term ideas (like new laws for new risks). Companies should follow self-rules, fix complaints, and use global standards (BIS/ISO). Regulators should use facts and stay flexible.