A new AI system can read a single face photo and give clues about a person’s biological age and cancer survival chances. The technology shows how a simple selfie might become a useful health indicator.

Researchers built a foundation model called FAHR‑Face. It was trained on more than 40 million facial images that cover a wide range of ages and ethnic backgrounds.

From FAHR‑Face they created two tools. FAHR‑FaceAge estimates a person’s biological age, while FAHR‑FaceSurvival predicts how likely a cancer patient is to survive, using only one photo.

The age tool looks mainly at the temples, the forehead above the eyes, and the nasolabial folds. The survival tool focuses on the bridge of the nose and the area under the eyes.

In a test, 1,060 patients whose estimated biological age was more than 20 years older than their real age saw a 21 % death rate within 90 days. By contrast, about 6 % of 4,195 patients estimated at least 10 years younger died in the same period.

The researchers also found that the models gave consistent results despite changes in pose, lighting, facial expression, makeup, or most plastic‑surgery procedures, with only a slight shift after rhinoplasty. This stability suggests the AI could work in real‑world settings.