“Although we understand that it’s not a 100 percent accurate technology yet,” said Tony Montoya, the president of the San Francisco Police Officers Association of facial recognition technology. “It’s still evolving,” he said, opposing a ban on San Francisco Police Department (SFPD) use of such gear.
The ban on police use of facial recognition technology in the city of San Francisco carried the City’s Board of Supervisors by an 8 to 1, and it lays down a clear marker.
But it’s a theoretical marker because the police department does not currently use facial recognition technology, and the places that do in San Francisco (the international airport and ports) are under federal jurisdiction, and thus not covered by the ban.
Still, similar laws are being considered across the bay in Oakland and across the country in Somerville, Massachusetts, a suburb of Boston, while the commonwealth of Massachusetts is considering a moratorium on facial recognition policing.
On the other hand, facial recognition technology is reported to be currently in use by police departments in Las Vegas, Orlando, San Jose, San Diego, New York City, Boston, Detroit and Durham, North Carolina.
So that makes the question how far below 100 percent is the accuracy of facial recognition systems pretty important. Dozens are on the market and many claim their matches of faces caught on camera and those in a database run in the high 90th percentiles.
But the FBI said last year its facial recognition system, and its enormous database of faces has an 85 percent chance of making a correct identification, and an even more recent MIT study of three leading private systems said when it came to dark-skinned women, the facial recognition machine had a less than two out of three chance of assessing gender, much less identity.
Still, Tony Montoya of the SFPD says, “I think it has been successful in at least providing leads to criminal investigators.”
“Police put their total faith in technology, expecting it to solve the massive, seemingly intractable problems inherent in their work,” says our guest today, investigative reporter Matt Stroud, in his new book Thin Blue Lie. Stroud considers such technological advances as Tasers, body cameras, surveillance cameras, cell phone trackers and such tech-based systems as Compstat and predictive policing and reports, “in many cases they don’t do what they’re supposed to do.”
Matt Stroud is an investigative reporter with a focus on companies that do business with police departments and prisons. Formerly on staff at the Associated Press, Bloomberg News, and The Verge, he has also written for publications such as The Atlantic, Politico, Buzzfeed, and The Intercept. Thin Blue Lie is his first book. Stroud lives in Pittsburgh, Pennsylvania.