India: India AI Governance Guidelines 2025 (MeitY; seven sutras, no new AI-specific statute)

1. Part 1 sets out the seven sutras that ground
2. Part 1: Key Principles
3. Part 2: Issues & Recommendations
4. 2.1 Infrastructure
5. 2.2 Capacity Building
6. 2.3 Policy & Regulation
7. 2.4 Risk Mitigation
8. 2.5 Accountability
9. 2.6 Institutions
10. Part 3: Action Plan
11. Part 4: Practical Guidelines for Industry & Regulators
12. Part 1 – Key Principles
13. 03 All other things being equal, responsible innovation should be prioritised
14. 06 Provide disclosures and explanations that can be understood by the
15. 07 Safe, secure, and robust systems that are able to withstand systemic
16. Part 2 – Key Recommendations:
17. 01 Infrastructure
18. 02 Capacity Building
19. 03 Policy & Regulation
20. 04 Risk Mitigation
21. 05 Accountability
22. 06 Institutions
23. Part 3 - Action Plan
24. Part 4 – Practical Guidelines
25. 01 Trust is the Foundation
26. 02 People First
27. 03 Innovation over Restraint
28. 04 Fairness and Equity
29. 06 Understandable by Design
30. 07 Safety, Resilience and Sustainability
31. 2.5 Accountability 31
32. Part 3: Action Plan 38
33. Part 4: Practical Guidelines for Industry 42
34. 01 Accountability
35. 03 Agentic AI responds to its environment and takes actions
36. 04 AI Incident
37. 05 AI Safety Institute
38. 06 Algorithmic Trading
39. 07 Artificial Intelligence
40. 08 Auditability The ability to inspect and verify system processes
41. 09 Behaviour Audit Evaluating AI decisions in real-world settings for
42. 10 Bias objects, people or groups in comparison to
43. 11 Data Minimisation
44. 12 Data Poisoning
45. 15 Explainability factors influencing the AI system results in a way
46. 16 Fairness
47. 17 Federated Learning node across a distributed network trains a global
48. 18 Foundation Models
49. 19 Generative AI Models that generate text, images, or other
50. 20 GPU (Graphics Processing and image processing, and specialized tasks in
51. 21 Human in the loop/ particularly during training and deployment to
52. 22 Large Language Models
53. 23 Machine Learning through computational techniques, such that
54. 24 Model Bias assumptions during the modelling process, that
55. 25 Red Teaming
56. 26 Small Language Models processing, understanding & generating natural
57. 28 Understandability
58. 01 IS/ISO/IEC 20546: 2019
59. 02 IS/ISO/IEC/TR 20547-1: 2020 architecture
60. Part 1: Framework and application process
61. 03 IS/ISO/IEC 20547-3: 2020 architecture
62. Part 3: Reference architecture
63. 04 IS/ISO/IEC 22989: 2022
64. 05 IS/ISO/IEC 23053: 2022
65. 06 IS/ISO/IEC 23894: 2023
66. 07 IS/ISO/IEC/TR 24028: 2020 Overview of trustworthiness in artificial
67. 08 IS/ISO/IEC/TR 24029-1: 2021 robustness of neural networks
68. Part 1: Overview
69. 09 IS/ISO/IEC 24029-2: 2023 robustness of neural networks
70. Part 2: Methodology for the use of formal methods
71. 10 IS/ISO/IEC/TR 24030: 2024
72. 11 IS/ISO/IEC/TR 24368: 2022
73. 12 IS/ISO/IEC/TR 24372: 2021 Overview of computational approaches for AI
74. 13 IS/ISO/IEC 24668: 2022 Process management framework for big data
75. 14 IS/ISO/IEC/TS 25058: 2024
76. 15 IS/ISO/IEC 25059: 2023 Quality Requirements and Evaluation SQuaRE
77. 16 IS/ISO/IEC 38507: 2022 Governance implications of the use of artificial
78. 17 IS/ISO/IEC 42001: 2023
79. 18 IS/ISO/IEC/TS 4213: 2022 Assessment of machine learning classification
80. 19 IS/ISO/IEC 5338: 2023
81. 20 IS/ISO/IEC 5339: 2024
82. 21 IS/ISO/IEC/TR 5469: 2024
83. 22 IS/ISO/IEC 8183: 2023
84. 23 IS/ISO/IEC 5259-1: 2024 and machine learning (ML) —
85. Part 1: Overview, terminology, and examples
86. 24 IS/ISO/IEC 5259-2: 2024 and machine learning (ML) —
87. Part 2: Data quality measures
88. 25 IS/ISO/IEC 5259-3: 2024
89. Part 3: Data quality management requirements
90. 26 IS/ISO/IEC 5259-4: 2024 and machine learning (ML) —
91. Part 4: Data quality process framework

The outline comes from headings detected in the text, so it might miss some or include extras.