NYPD not true blue when it comes to facial recognition transparency
The worst-case scenario for those who want police to use a controversial facial recognition service is playing out.
News is breaking nationwide of police agencies using Clearview AI‘s face biometric service reportedly with insufficient oversight, sometimes despite government assurances to the contrary and without the knowledge of top brass.
Occasionally it is happening in direct violation of explicit department policies or even government directives. Willful ignorance by the public about digital biometric face-matching programs in general could switch to opposition as a result.
Here is how. Muckrock, a collaborative government-transparency publisher, last week posted emails between the New York Police Department and Clearview AI that were divulged only after freedom of information requests.
Prior to the publication of the emails, department officials had been quoted saying the department had no relationship “formally or informally” with Clearview AI.
Yet the communications, spanning October 2018 to February 2020, detail an ongoing and broad relationship that is “well developed,” according to MIT Technology Review editors who analyzed the trail.
In fact, an investigation by online publisher BuzzFeed News looking at how 1,800 public entities have used Clearview AI has found that the NYPD has performed 11,000 searches using Clearview AI, including as part of live investigations.
MIT Technology Review, looking at the emails, found that total seemed to indicate a level of activity that would violate policy requiring that facial recognition be used only by a specific team. Indeed, “many officers outside the facial recognition team” used Clearview AI’s app, according to publication editors.
What is more, some officers had put the app on their personal devices, including their phones, something forbidden by the department.
New York was not the only jurisdiction examined in the wake of Buzzfeed’s reporting.
The (San Jose) Mercury News followed up and found that police officers in Alameda in the Bay Area have been using Clearview AI despite a ban on use of facial recognition tools signed in December 2019.
Some Idaho law enforcement agencies, according to Idaho Statesman, have tried the app, but, after limited use, have decided public opposition to it would outweigh the benefits.
Coincidentally, legislation designed to ban facial recognition has progressed in New Hampshire and Virginia. Both demonstrate a deep level of distrust in the technology and those wielding it.
The New Hampshire House of Representatives approved a measure outlawing the use of facial recognition by any state or local government agency unless a search warrant claiming probable cause is issued. The matter moves to the Senate now.
And Virginia legislators earlier this month acted to effectively ban the use of facial recognition by police.