by Dale Chappell

The Los Angeles Police Department (“LAPD”) denied for over a decade that it used facial recognition software. It even boldly said it never used facial recognition when responding to public records requests for data on facial recognition use. “We actually do not use facial recognition in the Department,” the LAPD told the Los Angeles Times in 2019.

But it has. A lot. LAPD now admits that it used facial recognition almost 30,000 times since 2009, with over 300 officers accessing a database of 9 million photos. But technically, the LAPD is correct: It didn’t really use facial recognition “in the Department.” Instead, it allowed its officers to feely access a facial recognition system in use by the L.A. County Sheriff’s Department. They have a direct link to it.

Now LAPD says that its misleading the public about using facial recognition was merely a “mistake.”

The bigger problem is with the software the county uses. It’s run by Dataworks, which made the news recently for leading to two false arrests based on facial recognition searches run on its system.

Now the LAPD is cutting the use of the controversial facial recognition software called Clearview AI, an unproven facial recognition platform that mines photos from the internet. This makes it a highly questionable resource for a government agency. Clearview AI has already lost most of its private customers over lawsuits.

Forbidding the use of Clearview AI by LAPD will be only a drop in the bucket, however. That’s because LAPD officers will still have unfettered access to the city’s facial recognition database, but at least its finally admitting what was known for years: That it uses facial recognition to pursue charges against citizens.

When the public misleads the police, it’s called obstruction. When the police mislead the public, it’s called a mistake.

Source: techdirt.com

Subscribe today

Already a subscriber? Login