The proposal out of the Senate is the latest in a string of efforts to limit the use of the technology in Utah. Allegation of misuse first surfaced in 2019, kicking off a long-running conversation about guardrails.
In response to privacy concerns regarding Utah’s use of facial recognition software by law enforcement agencies, the state’s Senate recently approved a bill to prevent law enforcement from scanning the drivers’ license database using facial recognition technology — unless they’re investigating high-level crimes.
In 2019, the Washington Post published an article about a report by Georgetown Law’s Center on Privacy and Technology stating that the Department of Public Safety (DPS) allegedly scanned driver’s license photos to search for a wanted person at the request of out-of-state interests, including U.S. Immigration and Customs Enforcement (ICE) and the FBI, without warrants.
Brian Redd, chief of the Utah Department of Public Safety and State Bureau of Investigation, disputed the allegations, saying, “we’ve always had a policy on using facial recognition.”
“The software is not used for immigration purposes or scanning driver licenses to that extent,” Redd said. “There is no way for someone to access the drivers’ license database.”
However, he said, the report and concerns from citizens ended up sparking an in-depth discussion about the issue.
At the beginning of the year, a bill sponsored by Utah Sen. Daniel Thatcher, R-District 12, was introduced outlining new guidelines for using facial recognition technology by law enforcement and government agencies.
The bill, titled Senate Bill 34 Governmental Use of Facial Recognition Technology, states:
Essentially, Redd said, the changes made include requiring that local law enforcement make requests through the Unified Criminal Justice Information System to run drivers’ licenses or someone’s criminal history or submit crime evidence.
“There has always been tracking and auditing of this information,” he said, “but there hasn’t been a criminal penalty before for the misuse of this system.”
As for other concerns related to using facial recognition tech, Redd said, “there have been concerns that the algorithm associated with this tech is less accurate in identifying people of color or females, so we have a policy in place that when the system checks or returns results that two FBI trained analysts work together to confirm the information.”
Another concern, he said, specifically from the American Civil Liberties Union, is that the tech isn’t used for misdemeanor crimes like jaywalking, for example.
“We have been using this system for over 10 years now and are always looking to protect the privacy of others,” he said. “We are listening to comments being made and are doing what is necessary to achieve this.”
The bill was moved to the House Jan. 2 for consideration.
Editor's note: An adjustment was made to identify the state agency at the center of the Georgetown Law’s Center on Privacy and Technology report.
Never miss a story with the daily Govtech Today Newsletter.