News

Manchester researchers have developed a systematic methodology to test whether AI can think logically in biomedical research, helping to ensure safer, more reliable applications in health care innovation. Their findings, published in an article posted to the arXiv preprint server, show that while AI can make intuitive connections, even advanced open-source models struggle with consistent logical reasoning when applied to biomedical problems. Danilo Carvalho, Principal Clinical Informatician for the Digital Cancer Research team at the National Biomarker Center, within Cancer Research UK Manchester Institute explains, By exposing where AI reasoning breaks down, we can build systems that support biomedical research with certain scientific evidence guarantees." The team's open-access methodology offers a vital tool for improving the transparency, reliability, and future design of AI technologies used in medicine, supporting Manchester's commitment to ensuring responsible AI and digital health innovation."