The groups say a Democratic-controlled government will be more receptive to their arguments of the software’s bias and privacy threats than the previous administration.
But the advocates are certain to face resistance from law enforcement and other facial recognition proponents who argue that a technology widely used to unlock phones and confirm travelers’ identities should also be made available to scan for wanted fugitives and investigate crimes.
The White House did not respond immediately to requests for comment.
The groups argue that the continued expansion of facial recognition runs counter to Biden’s inauguration-speech declaration that the government should work to advance “equity, civil rights [and] racial justice.” The technology has been shown in research to return inaccurate matches more often when assessing people of color, and it has been faulted in three separate wrongful arrests of Black men who were falsely identified by a police facial recognition search.
The technology’s developers and defenders argue that it is largely accurate and will improve over time. They caution that it should be used only as an investigative aid for officers looking for leads on whom to pursue, not as a primary piece of evidence.
But the letter’s signers argue that the software poses a clear threat to civil rights, regardless of its accuracy, because it could be used by government authorities to silently track innocent people or surveil protests.
“Even if the technology worked perfectly, it would facilitate the mass tracking of each person’s movements in public space — something intolerable in a free and open society,” the letter states. “We cannot allow its normalization.”
The face-scanning software can be used to identify people from afar without their knowledge or consent, and it works by quickly comparing a search photo against a vast database of driver’s licenses, mug shots and other images.
But federal investigators and police departments have also used the software for years to help identify potential suspects, victims and witnesses caught on surveillance cameras or seen on social media. Neither Congress nor state legislatures has authorized the development of such systems, and no clear federal oversight or regulations govern their use.
The technology has been banned or restricted by more than a dozen cities and states across the United States since 2019, including San Francisco, Boston, Portland, Ore., and Minneapolis. In December, New York became the first state to ban the technology’s use in schools until at least 2022, citing “serious and legitimate privacy concerns.”
But some cities have stuck by the systems. In Detroit, where the police chief said the system was useful even though it almost never returned a perfect match without human guidance, city leaders last year approved further use of the software, saying it helped protect the public while empowering the police. (Nearly all of the more than 100 people whose faces were run through a facial recognition search last year were Black, police data shows.)
The backers of the letter Tuesday urged Biden to sign an executive order halting federal facial recognition use “so long as bias pervades these systems and as long as Congress has not acted to authorize” its use.
They also pushed the White House to block state and local governments from using federal funds to buy or access the systems and to lend its support to a bill introduced last year by Sen. Edward J. Markey (D-Mass.) that would place a moratorium on federal deployment of facial recognition and other “biometric surveillance systems.”
The signers include Amnesty International, the Electronic Frontier Foundation, Freedom House and a number of other left-leaning activists, human-rights organizations and tech advocacy groups. But facial recognition has faced bipartisan criticism in Washington, including from congressional Republicans such as Rep. Jim Jordan (Ohio), who has said it poses dystopian threats of government overreach.
Local, state and federal law enforcement officials have revealed few details about how widely they use such software or how accurate the matches have been. A 2019 government audit found that the FBI’s facial recognition system, which analyzed more than 641 million photos from passports and other government databases, had been searched more than 390,000 times between 2011 and 2019.
The technology is also being increasingly deployed at U.S. travel hubs, with more than 23 million travelers scanned at 30 airports, seaports and land crossings last year, according to U.S. Customs and Border Protection officials, who called it “the way of the future” in a report last week.
Fewer than 300 “impostors” have been caught by the systems since 2018, the agency said. Critics argue the trade-off is not worth it, given that the mass data gathering can also backfire: Government auditors said last year that 184,000 images of travelers captured as part of a CBP facial recognition program were stolen by hackers during a 2019 breach of a CBP subcontractor, with some of them later posted to the dark Web.
CBP has promoted facial recognition as a quick and contactless way to confirm identification during the pandemic. And the Jan. 6 insurrection led a wave of federal investigators and amateur sleuths to use the software in hopes of identifying rioters who stormed the Capitol.
But Kate Ruane, senior legislative counsel with the ACLU, said she worries such uses may gloss over how the technology could further supercharge more surveillance in a “completely over-surveilled society,” where many already feel they are under close government watch.
“The power of anything that is this dangerous, that is this biased, to track us across space and time for unknown periods,” she said, poses a threat to Americans’ liberty “to go to a protest, to go to a church, to go to your doctor without the government surveilling you.”