East Bay State Sen. Steve Glazer, with his mortician-like visage, does not exactly fit the stereotypical image of an unsavory character, although, BART union members might disagree.

Glazer, who represents the 7th State Senate District, is among 26 members of the California state lawmakers who were mistakenly identified by facial-recognition software to have committed crimes in the past.

The American Civil Liberties Union used software created by Amazon, known as Rekognition, to match the legislature’s 120 members against a database of 25,000 mugshots. A similar study using members of Congress yielded 28 false matches.

The stunt is intended to underscore the problematic nature of facial recognition software, including its detrimental affect on communities of color, the ACLU said during a press conference Tuesday.

San Francisco Assemblymember Phil Ting is carrying legislation that would ban the use of the software in police body-worn cameras. No city in California currently uses the software in its body-worn cameras.

Advertisement