News Artificial Intelligence

Can fake faces make AI training more ethical?

Synthetic images may offer hope for training private, fair face recognition

Facial recognition surveillance interface showing multiple detected faces with green bounding boxes, tracking IDs, timestamps and frequency counts, displaying real-time identification of people in a crowd.

To achieve high accuracy, facial recognition models must be trained on millions of photos of people’s faces.

DAVID MCNEW/Contributor/Getty Images

By Celina Zhao

AI has long been guilty of systematic errors that discriminate against certain demographic groups. Facial recognition was once one of the worst offenders.