How India can move towards responsible AI in health

This article is authored by Dr. Ashok Puranik, executive director, AIIMS, Guwahati.

Every day in clinics and hospitals across India, a similar scene unfolds. Waiting rooms are brimming with patients, the administrative backlog is mounting, and each visit lasts shorter than the clinician would like. The strain of rising patient volumes is acutely felt by both doctors and nurses. This is where the promise of new-age, AI-powered tools to support clinical decision-making comes in. The solution to public health challenges and quality care is not simply more professional talent, but also building systems that empower clinicians and improve personalisation for patients.

AI

India is at a critical inflection point. Recent research by Elsevier found that clinicians in India are fatigued, with 79% seeing more patients now than they did two years ago. Out of those who feel that they are unable to deliver quality care, high patient volume emerged as the leading cause. This is a profession under pressure, where there is no scope for mistakes or missteps. Against this backdrop, the adoption of digital tools is increasing. Nearly 41% of clinicians have used an AI tool for work purposes, up from 26% in 2024. This is India’s moment to reimagine the clinical workflow, to preserve the human core of healing and make its medical systems stronger and more responsive.

Often, it’s the familiar, generalist chatbots like ChatGPT being co-opted for clinical tasks, such as summarising research, drafting patient communication, or clarifying complex concepts. Nearly half of clinicians globally (48%) now use an AI product for work. In India, the trend is clear: 36% have used clinical-specific AI tools. They perceive clinical AI tools as saving them time, empowering them, and giving them more choice in how they care for patients, helping them identify drug interactions and analysing medical images.

This bottom-up adoption speaks to the felt need for tools that ease clinicians’ burden and empowers them to deliver quality care. However, it also presents a silent risk: A fragmented and unguided integration of AI into the most sensitive of human domains—health. With only 12% of doctors reporting that their institutions provide good training on AI, India must proactively move the needle on standardised AI training for clinicians.

Regulation, while essential, moves at the speed of policy. Innovation moves at the speed of need. The gap between them is where risk flourishes. India cannot afford to wait for perfect laws to catch up. Instead, we must build a scaffolding of guidance, standards, and support that evolves alongside the technology, ensuring safety and efficacy today, not tomorrow. This can allow ‘safety-first’ AI tools to flourish—deliberate, validated, and clinician-vetted, which can enjoy a high degree of trust from clinicians and patients both.

India’s biggest gap is not technology; it is system readiness. India’s significant strides in Digital Public Infrastructure (DPI) show how systemic efforts can drive tech integration at the population scale. For the medical ecosystem, widespread adoption of clinical AI tools will be crucial to improving outcomes for doctors, nurses, and patients. The foundational need is is to improve access to digital decision-support tools for clinicians, offer robust guidance and training on AI usage, and create governance teams at organisational levels to manage AI policy and procurement.

This requires three distinct dimensions:

• Infrastructure & governance: Interoperable Electronic Health Records (EHRs) and secure, standardised data pipelines are the lifeblood of effective AI. Fragmented data leads to fragmented insights. Governance ensures proper auditing, reviewing outputs, and reducing ambiguity
• Training: It is unreasonable to expect clinicians to use a tool they do not understand. Training must move beyond the ‘how-to-click’ to the ‘how-to-interpret-and-integrate.’ and creating best practices that can be openly shared
• Culture: A culture of fear around AI will stifle it. A culture of blind faith will endanger patients. We must foster a culture of informed trust—critical, engaged, and participatory where clinicians take the lead in shaping the use of AI in health care.

Doctors must be at the forefront of change. 57% of clinicians globally believe AI will save them time. In India, where time is the scarcest clinical resource, that promise can translate into millions of additional meaningful patient interactions, reduced burnout, and better health outcomes. Policymakers, regulatory bodies, hospital boards, and institutions must work together to support clinicians in adopting AI tools to strengthen India’s medical system.

As the critical human bridge, it’s ultimately clinicians who translate AI’s output into compassionate care. This is possible when clinicians become advocates: demanding responsible AI in your workplace, asking the hard questions about validation, bias, and accountability, and pushing for inclusive design and better training. Clinicians lived experiences are the most valuable data for making AI tools effective and safe– this can create a feedback loop where AI tools continually improve and personalise according to the needs of patients. When AI tools offer admin and decision-making support, clinicians are able to focus on their best strengths: empathy, clinical judgment, and the therapeutic relationship.

The promise of AI in Indian health care will be realised not by the sophistication of the algorithms alone, but by the strength of the systems that support them and the trust of the clinicians who wield them. Let us build that foundation together, with responsibility as our compass, and the clinician-patient bond as our ultimate destination.

This article is authored by Dr. Ashok Puranik, executive director, AIIMS, Guwahati.