Handbook on Data Protection and Privacy for Developers of Artificial Intelligence (AI) in India
New Delhi: Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ); FAIR Forward – Artificial Intelligence for All (2026), 117 pp.
Contains acronyms pp. 7-9
"This Handbook is designed as a practical guide for developers of AI systems, especially from early-stage startups, to navigate data protection obligations and ethical considerations, in a clear and actionable manner. Rather than serving as a comprehensive legal or
technical manual, the Handbook complements existing global and domestic resources on data protection and responsible AI. It offers a context-aware framework grounded in core legal and ethical principles, encouraging developers, product teams, and founders to interpret and apply these principles in ways best suited to their technology and user base. Drawing from existing frameworks, the Handbook offers recommendations across the lifecycle of an AI system.
[...] The Handbook is divided into two main sections: Section I: Data Protection – which unpacks key concepts and compliance requirements under India’s data protection law, with a focus on their relevance to AI development; Section II: Responsible AI – which explores AI development through widely recognised responsible AI principles and provides a framework for their practical application. Each section concludes with a checklist of actionable takeaways intended to support developers in embedding privacy and ethical safeguards from the earliest stages of product design through to deployment. We have annexed a few case studies at the end, which demonstrate how developers adopt privacy and responsible AI principles in real-world applications." (How to read this handbook, page 10)
technical manual, the Handbook complements existing global and domestic resources on data protection and responsible AI. It offers a context-aware framework grounded in core legal and ethical principles, encouraging developers, product teams, and founders to interpret and apply these principles in ways best suited to their technology and user base. Drawing from existing frameworks, the Handbook offers recommendations across the lifecycle of an AI system.
[...] The Handbook is divided into two main sections: Section I: Data Protection – which unpacks key concepts and compliance requirements under India’s data protection law, with a focus on their relevance to AI development; Section II: Responsible AI – which explores AI development through widely recognised responsible AI principles and provides a framework for their practical application. Each section concludes with a checklist of actionable takeaways intended to support developers in embedding privacy and ethical safeguards from the earliest stages of product design through to deployment. We have annexed a few case studies at the end, which demonstrate how developers adopt privacy and responsible AI principles in real-world applications." (How to read this handbook, page 10)
SECTION I: DATA PROTECTION, 11
Summary of law and Key concepts, 13
Personal Data: Use for training AI models, 18
Data Sources, 26
Notice and consent, 30
Identifying a legal basis: Legitimate uses, 37
Understanding fiduciary-processor relationships in the AI lifecycle, 39
Individuals’ rights, 43
Organisational Measures, 47
SECTION II: RESPONSIBLE AI, 51
Fairness, 53
Transparency, 63
Accountability, 70
Security, 75
ANNEXURE – CASE STUDIES, 82
Case Study 1 : Cough Against TB tool, 82
Case Study 2: Prevention of Adverse TB Outcomes (PATO), 85
Case Study 3: Shishu Maapan, 87
Case Study 4: Krishi Saathi, 89
Case Study 5: Digital Green, 91
Summary of law and Key concepts, 13
Personal Data: Use for training AI models, 18
Data Sources, 26
Notice and consent, 30
Identifying a legal basis: Legitimate uses, 37
Understanding fiduciary-processor relationships in the AI lifecycle, 39
Individuals’ rights, 43
Organisational Measures, 47
SECTION II: RESPONSIBLE AI, 51
Fairness, 53
Transparency, 63
Accountability, 70
Security, 75
ANNEXURE – CASE STUDIES, 82
Case Study 1 : Cough Against TB tool, 82
Case Study 2: Prevention of Adverse TB Outcomes (PATO), 85
Case Study 3: Shishu Maapan, 87
Case Study 4: Krishi Saathi, 89
Case Study 5: Digital Green, 91