India is facing growing scrutiny over the cross-border flow of its sensitive health data, as global corporations increasingly look to low- and middle-income countries (LMICs) for training large AI models. Experts warn that without strong data protection frameworks and domestic investment in AI, India risks becoming merely a raw data supplier in the global digital health economy.
Rising Global Demand for Health Data
According to recent data, more than 5,000 peer-reviewed papers on AI in healthcare were published globally in 2023, up 1.5 times from the previous year. Much of this boom relies on training data drawn from LMICs like India, where patient data is more easily accessible due to weaker regulations and resource gaps in healthcare oversight.
With over £68 billion invested globally in healthcare AI over the past decade, the appetite for large, diverse datasets continues to grow. India’s high burden of non-communicable diseases (NCDs) and underfunded public health infrastructure make it an attractive source for clinical data.
How Health Data Leaves India
Foreign-funded research partnerships with charitable hospitals and clinical institutions in India often involve blanket consent agreements. These allow anonymised patient data to be collected, processed, and exported for proprietary AI development overseas. While legal, such practices raise ethical concerns over fairness and transparency.
Industry observers note that this dynamic is reminiscent of previous decades where clinical drug trials were offshored to LMICs to bypass strict oversight in developed countries. Today, the same logic applies to the harvesting of digital health data, including diagnostic scans and genetic profiles.
Underinvestment in Domestic AI Ecosystem
Despite the strategic value of health data, India allocated just ₹50 crore in its 2024–25 Union Budget for domestic AI research in healthcare. In contrast, countries in the Global North continue to invest billions into building proprietary health AI models—many trained on data sourced from the Global South.
According to the latest National Health Accounts, Indian households still bear over 45% of total health expenditures out-of-pocket, while government spending remains below 3% of GDP. This fiscal imbalance limits India’s capacity to build its own AI innovation pipeline.
Also read: The Cloud Revolution in Healthcare
Policy Recommendations: From Safeguarding to Sovereignty
To address these challenges, experts recommend a multi-pronged strategy:
Establish Trusted Research Environments (TREs): Secure, regulated platforms where Indian health data remains onshore, with access restricted to approved domestic users.
Create a National Health Data Commons: A public repository of anonymised health datasets for ethical, public-benefit research and AI development.
Mandate Benefit Sharing: Require foreign companies to share intellectual property or revenue if their AI tools are trained on Indian datasets.
Scale Domestic AI Models: Replicate successful local initiatives like Kerala’s AI-based diabetic retinopathy program nationwide.
Institutionalise Patient Involvement: Create public and patient groups to co-design AI tools, oversee data use, and ensure community benefits.
Conclusion
India’s vast medical data reservoir has immense potential to power next-generation healthcare solutions. However, without proper safeguards and investment, it risks reinforcing global inequities in innovation. Treating health data as a sovereign asset—on par with oil or minerals—will be key to ensuring that AI in healthcare serves national interests and strengthens public health outcomes.
