India’s contribution to global health is undisputed. From CSIR-IICT’s process chemistry for antiretrovirals and Cipla’s dollar-a-day HIV therapy, which reshaped the global response to AIDS, to its present leadership in generic medicines, India has earned the title “pharmacy of the world”. At the same time, we must accept that India has lagged in the high-value segment of new drug discovery.
An intermediate position is now emerging. India has entered biosimilars, has produced a handful of new drugs and promising drug leads, and its pharmaceutical companies are using their financial strength to acquire the intellectual assets of young international biotechnology firms. What ultimately matters for patients is the availability of high-impact products made in India for the world.
There are important parallels in another emerging field: the application of artificial intelligence to health. India has made progress on indigenous foundation models but remains well behind the leaders in the race to build frontier AI systems.
This should not constrain our health-AI ambitions. A “good-enough” system designed and formatted for the needs of a health system may outperform a frontier model deployed without such adaptation. A tuberculosis or diabetic-retinopathy screening tool, tuned to Indian data and wired into referral, is worth more to a district hospital than a frontier chatbot that is not. India has both an enormous need and the digital public infrastructure to deliver such solutions at population scale, and our solutions are likely to be relevant to much of the Global South, which faces similar constraints.
The most consequential opportunity is to use AI to shift healthcare from treatment towards prevention. Today, too much care begins after symptoms become severe, when treatment is more expensive and outcomes poorer. The question is no longer whether frontier models can beat specialists, but whether tools can anticipate illness, prioritise risk, intervene sooner, and improve chronic disease management. Even relatively simple systems, if clinically validated and integrated into care pathways, can generate substantial value. Prevention at scale will depend less on one spectacular algorithm than on thousands of reliable decisions embedded in everyday care.India’s digital public infrastructure is an exceptional foundation for this. Digital identity, interoperable payments, consent-based data exchange and the expanding Ayushman Bharat Digital Mission show that India can build common platforms at continental scale. Such infrastructure makes AI more effective by connecting it to real delivery systems. A screening tool is useful only if the at-risk person can be referred, tested, treated and followed up. A teleconsultation matters only when prescriptions, diagnostics and medicines reach the patient. Digital public infrastructure can join these pieces into one accessible care journey.None of this means India should abandon frontier ambitions. It should retain the scientific capacity to contribute to advances in model architecture, training and evaluation. But frontier ambition need not mean rebuilding every layer from the ground up. Much of the capability India needs already exists in increasingly powerful open-weight models, some of which approach frontier performance on important tasks. Indian generics did not win by inventing molecules. They won on process, scale and price. Open-weight models offer the same opening: a common technical base that Indian researchers and companies can inspect, adapt and improve. The higher-value work is to make these models reliable in Indian languages, with Indian clinical guidance, across uneven infrastructure, and under conditions where cost, privacy and continuity of service matter. This is also the route to real technological sovereignty. India should be able to choose, test, modify and, when necessary, replace the models on which essential services depend. Investment in frontier research and investment in open, locally adaptable systems are therefore complementary, not competing.
India should aim to build not merely an AI industry serving healthcare, but a global digital-health innovation ecosystem. This needs partnerships among hospitals, start-ups, pharmaceutical and technology companies, universities, public-health institutions and government. Two levers matter most. First, well-governed health-data sandboxes with clear access rules, so that responsible experimentation does not wait on case-by-case permission. Second, public procurement linked to ABDM that pays for validated tools beyond the pilot stage, so that credible innovations reach patients rather than conference posters. Regulation must demand evidence of safety, effectiveness, privacy and fairness.
India’s diversity is a further strength. Technologies that work across its languages, disease burdens, income levels and care settings will be relevant across much of the world. Just as Indian generics changed global access to medicines, India-designed health platforms and AI-enabled services could expand access to quality care throughout the Global South.
The strategic objective is not to win every race on terms set elsewhere. India need not build the world’s largest model to build the world’s most useful health-AI ecosystem. Our comparative advantage lies in combining appropriate technology, medical capability, entrepreneurship and population-scale public infrastructure.
A healthier Viksit Bharat will not emerge from technology alone. It will come from using technology to redesign healthcare around prevention, access and continuity of care. If India does this responsibly and inclusively, it can evolve from the pharmacy of the world into the health-AI laboratory of the world, creating solutions in India, for India and for humanity.
This article is written by Anurag Agrawal, Dean, BioSciences and Health Research, Trivedi School of Biosciences Head, Koita Center for Digital Health, Ashoka University, Delhi NCR
The views expressed are personal and do not necessarily reflect those of the university
(DISCLAIMER: The views expressed are solely of the author and ETHealthworld.com does not necessarily subscribe to it. ETHealthworld.com shall not be responsible for any damage caused to any person/organisation directly or indirectly)
