The stack, and why it changed the requirement
Until early 2026, an Indian health platform's compliance obligation was essentially about data: ABDM for interoperability and consent, the DPDP Act for protection. Artificial intelligence sat in a gap — governed by general principles, tested by whoever built it, validated against whatever data they happened to hold.
That gap closed on 17 February 2026, when the Ministry of Health and Family Welfare launched SAHI and BODH together at the India AI Impact Summit. SAHI is the policy framework; BODH is the platform that puts it into practice. Launching both at once — rather than writing a strategy and building its tooling years later — is unusual, and it means there is no grace period in which a health AI system can claim the tooling did not exist yet.
RISE will contain artificial intelligence in its predictive analytics, in clinical triage support, and later in air ambulance demand forecasting. All of it falls within scope — and because the system launches in 2028, it can be built to these frameworks rather than retrofitted to them, which is a genuine advantage of not having shipped yet.
Interoperability, consent, records
Ayushman Bharat Digital Mission. ABHA identity, the Health Facility and Healthcare Professional Registries, consent-based exchange and FHIR R4 profiles. The data rails everything else runs on.
AI governance, ethics, safety
Strategy for Artificial Intelligence in Healthcare for India. Five pillars, seven governing principles, thirty-two recommendations. Recommendatory, not statutory — and adopted here as binding on the programme.
Benchmarking and validation
Benchmarking Open Data Platform for Health AI, built by IIT Kanpur with the National Health Authority. Privacy-preserving evaluation of models against real anonymised Indian health data.
Cad Care RISE has not launched. This site states compliance intent and design, never achievement. Nothing here should be read as an assertion that the platform is currently certified, empanelled, audited or approved by any authority. Every claim carries its status in the conformance register, and any external version of this material must reproduce that register unaltered.
How the three fit together
The layering matters. ABDM makes the data available and consented. SAHI sets the terms on which an AI system may act on it. BODH provides the evidence that the system actually works on Indian patients rather than on the populations its training data came from.
A platform can satisfy any one of these and still be unsafe. ABDM compliance without SAHI governance produces a well-integrated system with no human oversight. SAHI adherence without BODH validation produces an ethically-designed system nobody has proven works. This is why the programme treats them as one obligation rather than three.