IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Opinion: California Wrong to Expand Facial Recognition Use

The technology still misidentifies individuals, especially when it’s focused on people of color. While the technology has advanced, the problems haven't gone away, and new legislation won’t fix them either.

California Capitol_shutterstock_1270097047
(TNS) — As technology advances, so too do the challenges surrounding the protection of the public's privacy and our civil liberties. That's particularly true regarding facial recognition software — the use of police body cameras to match people's photographs (from their driver's license or a photo taken at a traffic stop) and compare it to a database of mugshots.

Although facial-recognition technology offers advantages for police officials, it often is wildly inaccurate and can lead to the detention of people who have done nothing more than look someone else. A 2018 test of Amazon's software wrongly matched 28 members of Congress with suspects. The technology has advanced, but the problems haven't gone away.

The technology is particularly inaccurate when it comes to identifying people who aren't White, older people and children. In 2019, Assemblymember Phil Ting, D- San Francisco, authored a law that put a moratorium on police use of such technology until January of this year.

The Legislature needs to address the subject. It is considering two bills that take differing approaches.

This year, Ting has introduced Assembly Bill 642, which would require the state to "issue standards to ensure the confidentiality and cybersecurity of FRT data and results" and "require law enforcement agencies to establish written policies that adhere to those standards." Supporters say the bill would provide necessary safeguards, but ACLU California Action says it offers "a blank check for police and the surveillance industry."

The competing measure, Assembly Bill 1034, would extend the previous moratorium on police use of this technology only for body-worn cameras and police personal-use cameras until 2027. It would still allow police agencies to use the technology as part of its overall crime investigations. Not surprisingly, a wide range of civil-liberties groups support the bill and many police agencies and unions oppose it.

Although we appreciate that AB 642 tries to place limits on how police use the images, we agree with the ACLU. Once police have the green light to use this technology, it will become ubiquitous. The Assembly committee analysis notes that "this bill does not yet include all of the safeguards the author is hoping for."

During the AB 642 hearing, a Michigan man said he was arrested for theft based on a match between his license photo and surveillance footage, the Guardian reported. He noted in a letter that police were supposed to use the technology only as a lead but arrested him based on "an out-of-focus image of a large Black man ... that a faulty algorithm had determined was me."

By contrast, AB 1034's author Assemblymember Lori Wilson, D- Suisun City, explains that even facial recognition's proper use creates problems: "The widespread use of face recognition on police body cameras would be tantamount to requiring every Californian to show their photo ID card to every police officer they pass. Fear of mass police surveillance also could have a chilling effect on protests."

As critics note, facial-recognition software used by government agencies creates serious privacy problems, as individuals have no control over the use of the images. And it makes everyone susceptible to misidentification just by leaving the house. In an age of artificial intelligence, pranksters can spoof images. It gives government too much power to surveil us.

We support a continuing moratorium.

©2023 San Gabriel Valley Tribune, Distributed by Tribune Content Agency, LLC.