Demographic bias gaps are closing in face recognition, but how training images are sourced is becoming the field’s biggest privacy fight.

Imagine we are using a state-of-the-art face recognition algorithm and using a similarity value that falls into the -1 to +1 range.

The study published Thursday from the U.S. Department of Commerce's National Institute of Standards and Technology (NIST) looked at 189 software algorithms from 99 developers, which encompasses a ...