[Critically evaluate India's dual-track strategy of device-level regulation and ethical frameworks (SAHI) in governing AI in healthcare. (GS-III, 15 marks)](/upsc-mains-answer/critically-evaluate-india-s-dual-track-2bff7c2)

India governs health AI on two parallel tracks: a statutory device-level pathway, where AI-enabled medical devices are licensed under the Medical Devices Rules, 2017 by the CDSCO [1], and a policy-ethical track built on SAHI (Strategy for AI in Healthcare in India) and its benchmarking platform BODH, launched at the India AI Impact Summit 2026 [2]. The design is sound in principle, but its effectiveness turns on enforcement capacity.

Strengths of the dual-track approach

• Risk-proportionate legality: MDR 2017 already classifies devices into four risk classes, letting AI diagnostics be licensed without waiting for a dedicated AI statute [1].
• Evidence before deployment: BODH, developed by IIT Kanpur with the National Health Authority, evaluates models for performance, bias and generalisability on real-world data without sharing datasets — closing the "validated abroad, deployed in India" gap [2].
• Federal and ethical coherence: SAHI offers a common framework for Union, State and private actors [2], echoing WHO's 2021 principles of autonomy, accountability, transparency and equity [3].
• Global convergence: India's pioneer membership of the HealthAI Global Regulatory Network with the UK and Singapore aids harmonised standards [4].

Limitations

• A strategy is not a statute: SAHI carries no penalties; ethical duties remain non-justiciable.
• Regulatory gaps: MDR 2017 was framed for static hardware, and continuously learning or generative models — flagged by WHO in its 2024 LMM guidance for producing inaccurate or biased outputs — fit awkwardly into one-time licensing [5].
• Accountability vacuum: liability between developer, hospital and prescribing doctor is undefined, risking the outsourcing of clinical judgement.
• Equity risk: capacity and connectivity deficits may confine AI diagnostics to urban tertiary care.

The architecture is genuinely forward-looking — India is among the first states with a national health-AI strategy — but it is presently strong on vision and thin on enforcement. Statutory backing for post-market surveillance of adaptive algorithms, a clear liability rule anchoring final responsibility in the treating physician, and public-facility deployment of BODH-validated tools would convert the framework into assured, equitable and accountable care.