Global health summit calls for AI equity as diagnostics leap forward
WHA79 closes in Geneva; a multiomics blood test predicts heart disease 15 years early, and AI finds hidden GLP-1 drug risks.
By Dr. Asher Knippel
This week, from Geneva to Hong Kong, global health leadership and laboratory science converged — the 79th World Health Assembly closed with binding calls for health-AI equity, while researchers published a blood test capable of forecasting heart failure and stroke more than a decade in advance.
Monday, 25 May: World Health Assembly closes with teleradiology mandate and AI equity call
The 79th World Health Assembly (WHA79) concluded in Geneva with consequential resolutions shaping how artificial intelligence will be governed across health systems. Member States formally agreed to integrate teleradiology — the remote reading of X-rays and scans — into national health strategies, acknowledging AI's growing role in medical imaging. India's Union Health Minister J.P. Nadda warned that AI in health "could either bridge or deepen healthcare inequities depending on its design and regulation." The International Federation of Pharmaceutical Manufacturers & Associations (IFPMA) called at WHA79 for harmonised regulatory frameworks governing AI across drug development and pharmacovigilance, stressing the primacy of patient trust and data protection.
Saturday, 23 May: AI analysis of 400,000 posts uncovers hidden side effects of semaglutide and tirzepatide
University of Pennsylvania researchers trained an AI system on more than 400,000 Reddit posts from nearly 70,000 users taking GLP-1 receptor agonists — the drug class that includes semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound). The model identified side effects that rarely appear in clinical trial reports: menstrual irregularities, unexplained chills, and hot flashes. Findings were reported by ScienceDaily and corroborated by independent coverage. The authors suggest that AI-powered social-media pharmacovigilance could function as an early-warning system, capturing drug reactions that randomised trials — which are typically short, demographically narrow, and industry-funded — are structurally less likely to surface. Patients on these medications should raise any new symptoms with their prescribing clinician.
Saturday, 23 May: HKUMed blood test predicts six cardiovascular diseases up to 15 years before onset
A team at the Li Ka Shing Faculty of Medicine, University of Hong Kong (HKUMed), has published a multiomics cardiovascular risk tool called CardiOmicScore in Nature Communications. The single-blood-draw test analyses 2,920 circulating proteins and 168 metabolites, integrating genomics, metabolomics, and proteomics through deep learning to forecast individual risk of coronary artery disease, stroke, heart failure, atrial fibrillation, peripheral artery disease, and venous thromboembolism — up to 15 years before clinical onset. In validation cohorts, CardiOmicScore substantially outperformed conventional risk scoring. The tool is at the research stage and has not yet received regulatory clearance for routine clinical use.
Saturday, 23 May: UK study finds wide accuracy gaps among seven AI tools for lung cancer screening
A study published in the European Medical Journal evaluated seven CE-marked AI devices for detecting lung cancer on chest radiographs, using 5,235 scans from primary care patients in the United Kingdom. Diagnostic performance varied significantly across devices, raising patient-safety concerns as the NHS expands investment in imaging AI. The findings reinforce calls for standardised, real-world validation benchmarks before deploying AI screening tools at population scale — particularly in primary care, where patient demographics often differ from the hospital cohorts used to train most AI systems.
Sunday, 24 May: Anthropic and Gates Foundation commit $200 million to AI for global health
The Bill & Melinda Gates Foundation and Anthropic announced a four-year, $200 million partnership to develop AI tools for public health in low- and middle-income countries. Immediate priorities include AI-accelerated vaccine candidate screening — initially for polio — and research pathways for HPV-related disease and preeclampsia, a leading cause of maternal mortality worldwide. The partnership reflects a growing view that AI can compress research and development timelines in global health as substantially as it has begun to do in high-income pharmaceutical markets.
Monday, 25 May: Three senior physicians call for licensing clinical AI systems like doctors
Writing in the Journal of the American Medical Association, Alon Bergman, Robert Wachter (UCSF), and Ezekiel Emanuel (University of Pennsylvania) argue that the FDA's drug-and-device regulatory framework is inadequate for autonomous clinical AI. Their six-step proposal calls for federal competency certification, standardised AI examinations benchmarked against human clinical performance, and explicit accountability structures binding both AI developers and the institutions that deploy these systems. The opinion arrives as clinical AI products are being integrated into hospital workflows without the licensing oversight that governs the clinicians they work alongside.
Saturday, 23 May: MouseMapper reveals obesity's hidden damage to facial sensory nerves
Scientists at Helmholtz Munich and Ludwig Maximilian University Munich developed MouseMapper, an AI-powered system that maps an entire mouse body at cellular resolution. Using it, the team discovered that obesity damages facial sensory nerves associated with touch and sensation and triggers widespread neuroinflammation — effects previously undetected in whole-body surveys. When the researchers examined human tissue, they found comparable molecular patterns, suggesting the mechanism may translate to people. This is a preclinical study; all mechanistic findings derive from mouse models, and clinical follow-up studies will be needed before any conclusions can be drawn for human patients.
The content of this roundup is journalistic in nature and does not constitute medical advice. Readers should consult a qualified healthcare clinician before making any change to their treatment, medication, or health management plan.