The Alliance hosted experts in Montreux, Switzerland, at the end of September to explore how artificial intelligence (AI) can be responsibly and equitably integrated into health policy and systems research (HPSR), particularly in low- and middle-income country settings.
AI is already transforming health research, with applications ranging from reviews, synthesis and analysis of data to support with communicating and translating research. But HPSR is distinctive: it is equity-driven, participatory, and often embedded in policy processes. This makes the integration of AI both promising and complex. The consultation examined how AI can add value to HPSR and what safeguards are needed to ensure responsible, democratic and equitable use.
From capacity to evidence
Participants discussed the opportunities for AI to broaden capacity, lowering entry barriers for smaller teams and providing new avenues for mentorship, peer learning and multilingual collaboration. Dr Gauden Galea stressed how the open source community has empowered much of the progress in AI, and that, as we explore new developments for HPSR, we should utilize those approaches.
Evidence generation and synthesis was a central theme, with Gabriel Rada from Epistemonikos, sharing how systematic reviews are already being accelerated by AI tools.
“Systematic reviews were created to correct bias. If you just ask ChatGPT to summarize ten studies, the result isn’t reliable, because not all studies have the same weight. AI must be integrated into rigorous methods, not replace them”, he said.
Evidence-to-policy and governance
The meeting also explored how AI might support evidence-to-policy translation – from generating multilingual briefs and multimedia content to mapping stakeholder positions – while underscoring the continuing importance of human judgment and political negotiation. Cross-cutting concerns included governance, equity and data sovereignty, with calls for regional networks and shared frameworks to prevent dependency on a handful of global platforms. As several participants highlighted, AI should not drive the agenda, it should respond to what health systems and researchers actually need, based on equity considerations.
Launching new tools
The consultation coincided with the launch of the Alliance HPSR search tool – an AI-powered platform that allows users to quickly find and understand insights from Alliance-supported publications and journal articles. By linking AI directly with trusted HPSR sources, the tool aims to make evidence more accessible to researchers, policy-makers, and practitioners worldwide.
Looking forward
Kumanan Rasanathan, Executive Director of the Alliance, highlighted both the excitement and the responsibility that comes with AI’s rise: “We were excited to hear about how AI is changing research methods. But as we drive this work forward, we also want to find areas where HPSR itself can shape the development of AI tools – perhaps for more specific HPSR methods. AI holds lots of potential, but there are also many risks - our task is to manage these risks to ensure it strengthens health policy and systems research methods and approaches, taking particular care of its impact on equity.”
As AI becomes more embedded in research systems and health decision-making, communities and those affected by policy decisions should have a meaningful role in shaping how these tools are used and what values guide them.
The Montreux consultation marks an early step in a longer process. Insights from the meeting will inform a comprehensive report, to be released in 2026, on how AI can be deployed responsibly and equitably in HPSR to strengthen health systems and improve outcomes.