Fargo facial recognition saga sees police admit errors but refuse to close case on suspect

The twists and turns in the case could come from a Coen brothers script – but in the case of building public trust in police use of biometrics, that’s not a good thing. Fargo police have conceded they were wrong to arrest Tennessee woman Angela Lipps based on an erroneous facial recognition match, and will review their facial recognition policy as a result.

A report in Inforum says Fargo Police Chief Dave Zibolski has issued a temporary directive that instructs his staff to review the agency’s use of facial recognition to identify suspects.

Lipps was arrested at gunpoint in her home after Fargo police decided, counter to established laws and regulations, that a face match provided probable cause for an arrest. She spent four months in jail before bank records showed her being in Tennessee at the time of the alleged crime of bank fraud. Lipps alleges she was released from a North Dakota jail with no money and no coat in the middle of winter.

Oh, geez, we messed up: Fargo police

Zibolski says police are “more than happy to acknowledge our missteps along the way,” and are  “already taking actions to address that.”

The Fargo Police Department does not use its own facial recognition technology. Zibolski says  it sends photos for identification to North Dakota State and Local Intelligence Center (NDSLIC) and other places.

The technology used to identify Lipps was provided by Clearview AI and licensed to the West Fargo Police Department, which says it has received “numerous requests” from the Fargo force to use its facial recognition equipment. West Fargo’s system matched a photo from a fake ID used in one of their fraud cases to a photo of Lipps. But police opted not to send the match to prosecutors, and instead forwarded the data to Fargo detectives.

Zibolski, meanwhile, says the Fargo police didn’t know West Fargo police were using facial recognition in the Lipps case. “We would not have allowed that to be used, and it has since been prohibited,” he says. He also says his department should have sent the photos to the NDSLIC, per protocol.

After initially clapping back at allegations of wrongful arrest, Zibolski is now playing the aw-shucks card, shrugging off any notion of “malicious or intentional conduct.”

“I think maybe it’s a lack of knowledge or maybe an oversight in just covering all the bases in what they should have done to make sure we’ve got even more than what we should have had,” he says.

You’re darn tootin’ we got a criminal enterprise here

Lipps’ case was dismissed without prejudice, meaning the charges can be re-brought if more evidence comes to light. For his part, Zibolski seems determined to pursue her guilt in what he calls a “pretty organized criminal enterprise” training women to commit bank fraud through home equity withdrawals.

“If she is not involved in this in any way, we want nothing more than to be able to say that,” he said. “This is such an involved and complex scheme, and we’re just not certain. We want to be certain before we take that next step.”

Meanwhile, the chief’s facial recognition directive is already in effect, dictating that only the criminal investigation division at the Fargo Police Department can use facial recognition to identify suspects, and that cases must be reviewed by or “bona fide” state and federal facial recognition entities. Agents will receive training on biometric FRT, and all submissions will be tracked and reported monthly to the investigations division commander.

But Zibolski says there are no plans for Fargo police to contract their own facial recognition tool.

Meanwhile, West Fargo says it will continue to use Clearview’s tech, and to loan it out to other law enforcement agencies on request.

Article Topics

biometrics  |  Clearview AI  |  facial recognition  |  false arrest  |  law enforcement  |  North Dakota  |  police