AI Facial Analysis Predicts Cancer Mortality Risk, But Clinical Use Unproven

An AI system can identify cancer patients at high risk of dying soon by analyzing facial photographs, with over half of the highest-risk patients dying within 90 days of their picture being taken. The system, trained on over 40 million facial images, uses two tools to estimate biological age and predict survival. In tests, patients whose faces looked at least five years older than their actual age had a 21 percent higher death rate, and high-risk scores correlated with more than triple the death rate. The AI's clinical usefulness remains unproven, as it was tested mainly on White radiotherapy patients and cannot be applied to the general population. Researchers don't know exactly which facial features signal risk, and experts warn that errors could lead to patients being denied care or exploited by insurers.