Oakland City Council Votes to Ban Facial Recognition Software

Oakland’s City Council looks set to ban facial recognition software by city agencies after voting unanimously Tuesday night to approve the measure. A second vote in mid-September would finalize the ban, which would take effect immediately.

This makes Oakland the third city in the nation to do so after San Francisco and Somerville, Massachusetts. Berkeley is considering a ban as well.

The Oakland ban is similar to San Francisco’s, in part because activists like Brian Hofer, executive director of the nonprofit Secure Justice, helped craft both measures. He was in attendance at the City Council meeting.

Activists in other cities have talked to Hofer about doing something similar and in California.

“I hope we can build a national movement against dangerous face surveillance. We don’t need it,” Hofer said.

Facial recognition software uses machine learning algorithms to track human faces in digital footage and match them to names. But if the databases those software systems pull from are inaccurate or distorted by historic policing patterns, people of color can become the victim of false positives. Researchers argue that commercially available products for law enforcement aren’t able to correct for these systemic human biases, especially in criminal justice.

Oakland City Council President Rebecca Kaplan cited such concerns in a report written for the council. “Face recognition technology runs the risk of making Oakland residents less safe as the misidentification of individuals could lead to the misuse of force, false incarceration, and minority-based persecution.”

Law enforcement agencies have argued against outright bans, arguing that artificial intelligence has the capacity to improve with time and oversight, but they appear to be losing the political battle with a growing number of cities. The San Francisco ban, for instance, singled out facial recognition in a broader ordinance that required city departments to establish use policies and obtain board approval for all other kinds of surveillance technology they want to purchase or are using at present.

Among those testifying before the Oakland vote was Matt Cagle, an attorney with the Northern California chapter of the American Civil Liberties Union.

“You cannot leave your face at home,” Cagle said. “We do not want a world in which police are making decisions based on inaccurate information, misidentifying people and potentially using force based on flawed and inaccurate technology.”

The big software companies marketing facial recognition software are split on its use. In April, Microsoft refused to sell its technology to a California law enforcement agency over human rights concerns, and Microsoft President Brad Smith called for regulation to prevent “a commercial race to the bottom.”

Amazon, however, continues to sell its Rekognition software to law enforcement agencies. In a statement, the company argued, “Technology like Amazon Rekognition should only be used to narrow the field of potential matches. The responses from Amazon Rekognition allow officials to quickly get a set of potential faces for further human analysis. Given the seriousness of public safety use cases, human judgment is necessary to augment facial recognition, and facial recognition software should not be used autonomously.”

Regardless of the municipal bans, and a statewide moratorium on facial recognition for police body cams under consideration now in Sacramento, federal agencies like the FBI can and do use the software nationwide.

Facial recognition has been the subject of two House Oversight and Reform Committee hearings, in which concern over the technology appeared to be bipartisan.

Those hearings have yet to lead to a change in federal law, but they did reveal the FBI has amassed a database of more than 640 million photographs for its facial recognition program, including driver’s license photos from 21 states, not including California.

San Francisco Supervisor Aaron Peskin took that revelation as confirmation he made the right choice in voting for a ban. In an opinion piece published in Newsweek, he wrote, “Even if 100 percent accurate, mass surveillance by facial recognition — the likes of which authoritarian governments like China are already using — exposes Americans to an Orwellian dystopia.”