The proposal cleared the Senate earlier this week. It would only allow police to use facial-recognition tech when investigating a specific criminal incident or citizen-welfare situation.
Sen. Scott Surovell, D-Fairfax, said evidence gathered from facial-recognition tech could only be used for exoneration, not for establishing probable cause for an arrest.
“You can’t use it for broad surveillance or monitoring,” Surovell asserted. “You have to have a specific case you’re looking at, or you have to have a person in a hospital bed, and you don’t know who they are, and you’re trying to figure out who’s there, or you have a dead body, and you’re trying to figure out who that was and there’s no ID on them or whatever.”
Last February, the General Assembly passed a bill barring police from using facial-recognition technology unless they receive prior legislative approval, a measure The Associated Press referred to as “one of the most restrictive bans in the country.” Opponents of facial-recognition tech, including many legislative Republicans, argue it’s an invasion of privacy and prone to inaccuracy and abuse.
A 2019 report from the National Institute of Standards and Technology found Asian and Black people are far more likely to be misidentified by facial-recognition technology. The bill would require any facial-recognition tech used by police to be at least 98% accurate across all demographic groups.
Sen. Ryan McDougle, R-Hanover, expressed concerns on the Senate floor Tuesday the tech could still be misused.
“In this bill, even with the policies and the restrictions in place, there are no penalties if you violate it,” McDougle pointed out.
The measure also would require departments to log inquiries into their facial-recognition software, and then publish a public usage report at the end of each year. With its passage in the Senate, the bill now heads to the House and its committees for further deliberation.