AI in Public Health: Bridging Gaps from Pilots to Sustainable Impact

Abstract

The panel examined why AI projects in health often stall after the pilot stage and explored concrete pathways to embed AI‑driven solutions into India’s public‑health infrastructure. Drawing on perspectives from investors, government officials, European partners, philanthropists, CSR leaders and a mental‑health technology founder, the discussion highlighted diagnostic and screening use‑cases, data‑ownership and regulatory hurdles, the need for patient‑capital and ecosystem‑wide collaboration, and cultural‑context challenges that affect scaling. The session concluded with a set of actionable recommendations for government, funders, innovators and civil‑society to move AI from isolated pilots to durable public‑health assets for the Global South.

Detailed Summary

• Welcome and context – Sikka thanked participants for joining despite logistical challenges and positioned the conversation within the “People and Progress Sutras” of the India AI Impact Summit 2026.
• Core question – He asked CK Mishra to outline the Indian government’s view on scaling AI in health, emphasising the “missing middle” between promising pilots and nationwide adoption.

2. Government Perspective – CK Mishra

• AI as an enabler, not a panacea – Mishra stressed that AI is the most powerful technology the Indian health sector has seen, but it does not by itself deliver health‑care services.
• Strategic questions for any AI intervention –
  1. Access – Does it expand reach to underserved populations?
  2. Quality – Does it improve diagnostic or therapeutic quality?
  3. Equity – Does it reduce or risk widening health inequities?
  4. Affordability – Is it cost‑effective for low‑resource settings?
  5. Acceptability – Do health‑workers and patients understand and trust the tool?
• Diagnostics & screening as immediate impact zones – AI‑driven imaging, pathology and disease‑risk scoring can fill diagnostic gaps in Tier‑2/3 districts where specialist radiologists are scarce.
• Screening & early‑warning systems – AI can act as an “alert system” that reduces unnecessary hospital visits, improves preventive care and enables community‑level disease surveillance (e.g., obesity, hypertension).
• Primary‑care gap – Tele‑medicine platforms such as e‑Sanjeevani have limited diagnostic depth; Mishra argued that human‑in‑the‑loop remains essential, particularly for complex decision‑making.
• Capacity‑building need – Front‑line health workers often lack the digital literacy to operate AI tools; therefore awareness and training must precede technology rollout.
• Policy alignment – AI solutions must be embedded within existing national programs (e.g., National Health Mission) rather than exist as stand‑alone pilots.

3. European Union View – Pierrick Fillion Ashida

• Historical perspective – AI concepts (expert systems) have existed for decades; today the scale of data and compute power fuel rapid adoption.
• Data sovereignty & regulation – Ashida highlighted that ownership of health data belongs to the data subjects and emphasized the need for robust data‑protection legislation in India, mirroring EU GDPR principles.
• Complementary strengths – EU faces a doctor shortage; India faces a volume challenge. Collaboration could focus on:
  • Molecule design & drug discovery – leveraging AI for rapid prototyping.
  • Large‑scale screening – applying AI to rare‑disease detection given India’s population size.
• Interoperability – Standardized health‑record formats would enable cross‑border continuity of care for Indian patients traveling abroad.
• Open‑source advocacy – Both EU and Indian stakeholders should champion open‑source AI frameworks to avoid vendor lock‑in and to foster community‑driven innovation.
• Funding & partnership model – Suggests joint Gov‑to‑Gov pilots, VC‑to‑VC co‑funding, and fund‑of‑funds mechanisms to pool risk‑capital for scalable solutions.

4. Philanthropic Lens – Ishita Jain (CIF Foundation)

• Core mission unchanged – CIF remains focused on women and children, evidence‑based interventions, and alignment with government priorities.
• AI‑related equity risk – New AI tools can unintentionally reproduce existing health inequities if trained on non‑representative data. CIF will stress‑test models for bias, low‑bandwidth operability, and low‑literacy user interfaces.
• Accelerated but rigorous cycles – Philanthropic capital must be patient yet fast‑acting, allowing innovators to iterate quickly while maintaining scientific rigor.
• Funding mix – CIF plans to combine grant funding (for early‑stage R&D) with patient equity capital to bridge the “valley of death” between pilot and scale.

5. CSR Perspective – Nitin Vashisht (IOCL)

• CSR as “gap‑funding” – IOCL’s CSR budget (~₹500 crore annually) can only partially finance AI projects due to statutory compliance and short‑term (3‑year) project cycles.
• Capex‑focused approach – IOCL mainly supports equipment procurement (e.g., AI‑enabled TB diagnostic machines) for both premier institutes and rural health centers.
• Sustainability challenge – Capital‑intensive equipment needs operational‑expenditure (OpEx) models for long‑term impact, which CSR rules currently limit.
• Collaborative model – IOCL is forming hybrid partnerships with AI innovators to share data and expertise, extending the value of its capex beyond the CSR horizon.
• Policy bottleneck – Current CSR regulations restrict multi‑year funding, hindering continuous AI solution refinement and impact measurement.

6. Investment Landscape – Ajay Mahipal (HealthKois)

• Track record of scaling – HealthKois has backed several AI health‑tech firms that moved from Indian pilots to global deployments in >100 countries (e.g., VISA – mental‑health platform; Cure.ai – AI‑radiology).
• Three pillars for a successful AI health venture
  1. Intellectual Property (IP) & talent – Long‑term patient capital needed for algorithm development and talent retention.
  2. Pilots & adoption – Multiple, rigorously evaluated pilots are essential before commercial launch, especially in a regulated sector.
  3. Commercialisation – Robust business models, regulatory approvals (e.g., FDA/US‑FDA, WHO pre‑qualification) and market access strategies.
• Capital requirements – Scaling AI solutions across India’s massive health‑system (≈ 1 million doctors, 2 million nurses, 3 million beds) could demand > US$ 100 bn in capex and 10‑year infrastructure investment.
• Funding sources – Government grants, CSR, patient‑capital, venture capital, and deep‑tech funds (e.g., RBI‑backed RDIF Fund) are converging to create a more supportive financing ecosystem.

7. Startup Perspective – Namrata Mayanil (VISA)

• Cultural‑economic diversity – Adapting a mental‑health AI platform for Indian adolescents required ground‑level ethnographic research to align digital interventions with daily life (e.g., “DreamKit” that blends physical school activities with the app).
• Reimbursement pathways – Unlike the UK NHS or US insurers, India lacks a centralised payment mechanism for preventive mental‑health services, limiting commercial viability.
• User‑pay vs. insurer‑pay – Current Indian ecosystem places the cost burden on patients, employers or insurers, none of which currently prioritize preventive mental‑health spending.

8. Cross‑cutting Themes & Recommendations (Round‑Table Synthesis)

8.1. Systemic “Magic‑Wand” Recommendations (Speaker: Himanshu Sikka)

1. Stabilise, finance and promote the most suitable solutions – Provide long‑term, patient capital and policy incentives for proven pilots.
2. Integrate AI tools within national programs – Embed AI in existing schemes (e.g., National Digital Health Mission) to lower adoption barriers.
3. Design with end‑users in mind – Solutions must address real, locally‑identified health needs rather than being technology‑first.

8.2. Talent & Ethics (Speaker: Pierrick Fillion Ashida)

• Talent development – Joint EU‑India curricula on AI for clinicians, data scientists and health‑system managers.
• Ethical frameworks – Collaborative master programs on AI ethics, data ownership, and responsible deployment.

8.3. Philanthropy & De‑risking (Speaker: Ishita Jain)

• Incentivise innovators to align with government priorities.
• De‑risk early‑stage AI solutions through grant funding and pilot‑scale support, enabling later entry of private and CSR capital.

8.4. CSR Evolution (Speaker: Nitin Vashisht)

• Shift from output‑based to outcome‑based CSR budgeting, allowing longer‑term impact measurement.
• Encourage public‑private collaborations that extend beyond the 3‑year CSR cycle.

8.5. Investment & Market Maturity (Speaker: Ajay Mahipal)

• Expand deep‑tech funds (e.g., the INR 2,000 crore RDIF allocation) to nurture patient capital.
• Strengthen exit pathways (strategic acquisitions, IPOs) to attract more VC interest.

8.6. Cultural & Reimbursement Barriers (Speaker: Namrata Mayanil)

• Develop contextualized user experiences for diverse linguistic, socioeconomic, and literacy profiles.
• Advocate for a national mental‑health reimbursement framework (e.g., inclusion in Ayushman Bharat) to create sustainable demand.

Key Takeaways

• AI is a powerful enabler but not a substitute for health‑system fundamentals – policy, workforce capacity and financing remain decisive.
• Diagnostic and screening applications (AI‑assisted imaging, risk‑prediction tools) present the most immediate, scalable impact in Tier‑2/3 India.
• Data sovereignty and interoperability are prerequisite for cross‑border collaboration; India must adopt robust data‑protection laws and common health‑record standards.
• Long‑term patient capital (philanthropy, deep‑tech funds, CSR with outcome‑based metrics) is essential to move beyond pilots.
• Cultural and socioeconomic diversity demands locally‑tailored user experiences; one‑size‑fits‑all solutions fail at scale.
• Public–private partnerships should be institutionalised through government‑led sandboxes, joint funding mechanisms, and shared talent pipelines (EU‑India curricula).
• Regulatory reform (fast‑track approvals, clear AI‑device classification) and outcome‑based CSR can unlock sustainable financing.
• Ethics and equity must be baked into AI development; bias audits, low‑bandwidth design, and community‑led validation are non‑negotiable.
• Integration with national programs (e.g., NDHM, Ayushman Bharat) is the fastest route for adoption, avoiding siloed pilots.
• Future outlook: With emerging deep‑tech funds, growing VC interest, and increasing EU‑India collaborations, the next 5‑10 years could see a paradigm shift from isolated AI pilots to a cohesive, AI‑enhanced public‑health ecosystem across the Global South.

---

See Also:

• founders-funders-the-india-ai-capital-ecosystem
• ai-innovators-exchange-accelerating-innovation-through-startup-and-industry-synergy
• responsible-ai-for-bharat-building-trust-safety-and-global-leadership
• ai-for-everyone-empowering-people-businesses-and-society
• building-indias-ai-governance-architecture-from-frameworks-to-implementation
• designing-health-ai-for-bharat-a-billion-clinical-realities
• ai-impact-forum-democratising-ai-resources
• ai-for-industries-resilience-innovation-and-efficiency