Global AI Day: Reda Sadki on artificial intelligence for health #AI4Health

What if the biggest risk of AI in global health is not that it fails, but that it perfectly preserves the inequalities we have yet to flush out? In this talk at Global AI Day in Geneva on 6 May 2026, Reda Sadki, founder of The Geneva Learning Foundation, shares the AI4Health framework for treating artificial intelligence as a co-worker rather than a tool, and explains why peer learning networks between humans and AI agents may be the highest-return investment for health in low and middle income countries. Drawing on Eric Schmidt's San Francisco Consensus, Reda maps the three converging revolutions reshaping work: language as the universal interface, reasoning with working memory, and agentic workflows. He then confronts the new AI divide, where high-income knowledge workers gain specialized agents while frontline practitioners are left with basic chatbots, geo-locked tools, and the risk of what he calls epistemic injustice that kills. Learn about the AI4Health programme and access the full framework: https://www.learning.foundation/ai CHAPTERS 00:00 Introduction: human agency and the co-worker protocol 00:30 The San Francisco Consensus and the acceleration vector 01:00 The interface, reasoning, and agentic revolutions 01:30 Meet Claude Cardot, TGLF's first agentic AI co-worker 02:00 The new AI divide and the compute squeeze 02:30 Epistemic injustice: why frontline tacit knowledge disappears 03:30 From tool to partner: the architectural shift 04:00 The seven principles of AI partnership 04:30 Bridging global data and local reality: the market day example 05:30 The full learning cycle: analysis, action, mobilization 07:00 The multiplication effect: 7x faster, 90% lower cost 07:30 Applied: immunization and zero-dose children 08:00 Applied: climate change and distributed sensing 08:30 Applied: neglected tropical diseases and lateral knowledge transfer 09:00 Poverty porn 2.0 and the visual ethics filter 10:30 The transparency paradox in the humanitarian sector 11:30 The e-bike effect and the five Cs against metacognitive laziness 12:30 Investment roadmap for the new epoch 13:30 The strategic divergence: digital colonialism or networked intelligence KEY IDEAS - AI as co-worker, not tool - Epistemic injustice and the disconnect between frontline tacit knowledge and global norms - Peer learning networks that include both human and AI nodes - The five Cs: creativity, curiosity, critical thinking, consciousness, collective care - The danger that AI perfectly preserves the old poison we have yet to flush out ABOUT THE GENEVA LEARNING FOUNDATION The Geneva Learning Foundation supports a network of more than 80,000 humanitarian and health workers through peer learning for local action. The AI4Health programme explores how artificial intelligence can strengthen, rather than diminish, human agency at the frontline. Explore AI4Health: https://www.learning.foundation/ai #AI4Health #GlobalHealth #PeerLearning #ArtificialIntelligence #GenevaLearningFoundation #AgenticAI #DigitalHealth #HealthEquity #GlobalAIDay