Every AI model is flunking medicine - and LMArena proposes a fix

AI models from major companies are failing to provide safe and accurate medical information, according to a study by DataTecnica and the NIH's Center for Alzheimer's. Using the CARDBiomedBench benchmark, researchers found that no current model meets the knowledge demands for biomedical research. This is concerning as people increasingly trust AI medical advice. To address this, DataTecnica and LMArena.ai are expanding BiomedArena, a leaderboard evaluating AI models on medical research topics, aiming to develop tools that better serve the medical community.