Blog

Jun 23, 2023

Opinion: How automatic licensee plate readers make us less safe

Irani is an associate professor at UC San Diego and a member of the steering committee of TRUST Coalition. She lives in University Heights. Yusufi is deputy director of the Partnership for Advancement

Irani is an associate professor at UC San Diego and a member of the steering committee of TRUST Coalition. She lives in University Heights. Yusufi is deputy director of the Partnership for Advancement of New Americans and a member of the steering committee of TRUST Coalition. She lives in San Marcos.

Are automatic license plate readers ungovernable?

Police departments across the region want to adopt automated license plate reader technologies with the promise of solving crimes. These readers tie cars to locations, noting not only license plate numbers, but make, model, and features like bumper stickers and roof racks, according to a demo video by San Diego’s proposed vendor Flock Safety. Crucially, the devices allow police to connect plates to personal identity through state databases. (We challenge Flock’s branding as “safety” when scientific studies find automated license plate readers make little to no discernible impact on auto thefts.)

Commentary

In 11 days, El Cajon Police Department recovered 9 stolen vehicles and arrested 14 individuals.

Aug. 30, 2023

While we’re told that these technologies are tools for public safety, they actually make many of us less safe. Automatic license plate reader data can place a visitor from out-of-state near an abortion clinic or a clinic that provides gender-affirming care, and prosecutors in their state can request the data from San Diego. Federal immigration authorities have spent millions of dollars to buy automatic license plate reader data collected from cities like ours, using that data to locate our neighbors and separate their families. If you don’t want a Texas prosecutor in your neighborhood, you don’t want automatic license plate readers. If you don’t want Immigration and Customs Enforcement in your neighborhood, you don’t want automatic license plate readers.

Over the last year, TRUST Coalition members and many others have shown up to community meetings, Privacy Advisory Board meetings and City Council meetings, pressing our leaders to reject Mayor Todd Gloria’s automatic license plate reader proposal. Council voted yes on automatic license plate readers instead, but members adopted legal guardrails that attempt to prohibit using automatic license plate readers for immigration enforcement, abortion care prosecutions or minor crimes. We are heartened that our council members care.

These legal guardrails, however, are not guardrails at all. In practice, they are treated as little more than suggestions, and often ignored. We’ve learned from experience in San Diego that guardrail policies are commonly violated by those they are meant to constrain.

The California Values Act, or Senate Bill 54, offers a cautionary tale. SB 54 was passed by lawmakers in 2017 to protect immigrants by limiting local police and the use of state resources in assisting in deportations and family separation. Despite this, Voice of San Diego caught the San Diego Police Department in 2018 sharing its automated license plate recognition data with Customs and Border Protection and 600 other agencies throughout the country through its technology provider, Vigilant. In 2020, reporters revealed that the city of Chula Vista had shared its automated license plate recognition data with federal immigration officials for over three years, seemingly without the knowledge of its City Council.

California state law offers another reason for concern. Our state has had laws governing automatic license plate readers since 2015 requiring agencies to create privacy policies, keep systems secure and to notify us when our data is breached, as with any other personal information. So how did that law do in getting agencies to secure the automatic license plate reader data they collected?

Not great. A 2019 California State Auditor report looked at four local police departments’ automatic license plate reader practices. It found that none of the agencies it reviewed had legally compliant privacy policies. Three of four did not clearly define who could access the data. And three stored their data with an outside “cloud” data company without guarantees that the company would appropriately protect the data. More recently, Electronic Frontier Foundation caught El Cajon breaking the law by sharing automatic license plate reader data with Houston, where most abortions are illegal. California law offered guardrails and the local police blasted right through them.

The best way to keep data safe is to not collect it all. When data collection is necessary, European data law calls for “data minimization” — collect minimal data for specific purposes and delete it as soon as it is no longer required. Automatic license plate reader data minimization would mean, most generously, that automatic license plate reader data is only stored for incidents covered by a warrant. It does not mean what the City Council’s new policy allows — the dragnet collection and storage of surveillance for law enforcement fishing expeditions.

These technologies create an immense oversight burden for our public officials. In Chula Vista and in San Diego, we’ve seen City Council members surprised to learn at what agencies have done with surveillance. These technologies require complex collaborations between public agencies, technology companies and data storage companies that are hard to understand and even harder to track across states. Even worse, recent layoffs leave San Diego without as many reporters and researchers to help make sense of it all and keep government accountable.

Legal guardrails are meant to make us feel safe, but our own history shows that meaningful accountability is hard to come by and often too late. Data minimization and foregoing systems that make us vulnerable are the only pathways to make sure public safety for some does not become public danger for others.