Artificial Intelligence

AI Face Scanner Predicts Cancer Survival From Selfies

MAY 13 2025·Updated MAY 18 2025

Story

Above: The corporate offices of Massachusetts General Brigham hospitals in Assembly Square in Somerville, Mass. on Jan. 27, 2022. Image credit: Lane Turner/Contributor/The Boston Globe via Getty Images

The Facts

In research published in The Lancet Digital Health this month, it was revealed that Mass General Brigham researchers have developed an artificial intelligence tool, dubbed FaceAge, that analyzes facial photographs to predict biological age and survival outcomes in cancer patients.
Mass General Brigham
The study — which included 58,851 photos of presumed healthy people and 6,196 cancer patients — found that cancer patients appeared approximately five years older than their chronological age. Higher FaceAge estimates are strongly correlated to worse survival outcomes across multiple cancer types, even after adjusting for multiple factors.
Mass General Brigham
When clinicians used FaceAge data alongside traditional assessments, their accuracy in predicting six-month survival for palliative care patients improved from 61% to 80% — demonstrating the tool's potential value in treatment planning.
www.ft.com
The Washington Post
The AI system appears to weigh aging markers differently than humans do, focusing less on obvious signs like gray hair or baldness and more on subtle features, including facial muscle tone and temple hollowing, which may indicate overall health.
ScienceAlert
The New York Times
Researchers are now expanding testing across different hospitals and patient populations, while addressing concerns about potential biases in the training data — which primarily included white faces — and variables like makeup and lighting.
EuroNews
The New York Times

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Pro-establishment narrative

FaceAge is a groundbreaking advancement in precision medicine, offering doctors an objective, quantitative tool to enhance critical treatment decisions. The technology transforms subjective visual assessments into reliable biomarkers, potentially saving lives by helping clinicians better match treatment intensity to patient resilience.

Mass General Brigham

The Washington Post

Establishment-critical narrative

The rush to implement AI-based appearance analysis in health care not only raises privacy concerns but also risks perpetuating biases and creating new forms of discrimination. The technology's training data lacks diversity, and without proper safeguards, insurance companies could misuse this tool. Patients might also face psychological harm from learning they look older than their age.