Scientists have unveiled a groundbreaking artificial intelligence system capable of predicting an individual’s risk of developing more than 1,000 diseases, potentially forecasting health outcomes a decade or more into the future.
The tool, named Delphi-2M, was developed by experts from the European Molecular Biology Laboratory (EMBL), the German Cancer Research Centre, and the University of Copenhagen. Built using generative AI concepts similar to large language models, Delphi-2M represents one of the most ambitious demonstrations of AI applied to human health. Details of the research were recently published in Nature.
At its core, Delphi-2M analyzes “medical events” in a patient’s history—such as previous diagnoses—alongside lifestyle factors like smoking, alcohol use, obesity, age, and sex. By combining these data points with anonymized patient records, the system forecasts when and whether a person may face conditions ranging from cancer and diabetes to heart and respiratory diseases.
The model was trained on two vast datasets: the UK Biobank, containing information from 400,000 individuals, and Denmark’s national patient registry, with records from 1.9 million people. This scale allowed the system to identify patterns in disease progression across populations, generating forecasts with accuracy comparable to existing single-disease prediction tools. Unlike traditional models such as Qrisk, which focus on one condition, Delphi-2M evaluates risk for all major diseases simultaneously and over longer timespans—up to 20 years.
“Health conditions usually unfold in patterns we can recognize and learn from,” said Tomas Fitzgerald, staff scientist at EMBL’s European Bioinformatics Institute.
Researchers believe the tool could soon become a vital part of routine healthcare. “Imagine visiting your doctor and being shown your four biggest future health risks, along with clear steps to reduce them,” said EMBL interim director Ewan Birney.
The potential applications go beyond prevention. By generating synthetic “future health trajectories,” Delphi-2M could help doctors anticipate patient needs and tailor care strategies long before symptoms appear.
Prof. Moritz Gerstung, head of AI in oncology at the German Cancer Research Centre, emphasized the broader significance: “This is the beginning of a new way to understand human health and disease progression. Generative models like ours could one day personalize care at scale.”
