Virginia will lift its ban on the use of facial recognition by local police in July, while California and the city of New Orleans will “press the undo button” as soon as this month.
In New Orleans, where homicides have risen 67 percent over the past two years, police say they need every piece of equipment that works.
“Solving these crimes requires technology and holding people accountable,” Police Superintendent Shaun Ferguson told reporters, calling on the city council to repeal a ban that went into effect last year.
Efforts to get bans in place are running into resistance in jurisdictions large and small, from New York and Colorado to West Lafayette, Indiana. Even Vermont, the last state to have a nearly 100 percent ban on police use of facial recognition technology, cut content in its legislation last year to allow investigations of teenage sex crimes.
From 2019 to 2021, roughly two dozen U.S. state or local governments have enacted legal guidelines restricting facial recognition. Research has found that the technology is far less effective at identifying black people, and the anti-police “Black Lives Matter” protests have fueled that debate.
However, ongoing analysis by the federal government’s National Institute of Standards and Technology has demonstrated important advances in accuracy across the industry. The Department of Homeland Security found in tests last month that there was little difference in accuracy between people of different skin colors and genders.
“More and more people are curious about coverage methods for dealing with issues related to technology, while ensuring that it is exploited in a bounded, appropriate and non-discriminatory way that benefits the community,” lobby group Security Industry Association said Jake Parker, senior director of chairman relations.
The shift in sentiment could allow its members, including Clearview AI, Idemia and Motorola Solutions, to capture a higher share of the $124 billion that state and local governments spend annually on policing. The part dedicated to technology is not closely tracked.
This week, Clearview settled a privacy lawsuit over its collection of photos from social media by agreeing not to promote its flagship system to the U.S. personal sector.
Clearview, which helps police find matches in social media messages, said it welcomes “any regulations that help society get the best possible benefit from facial recognition technology while limiting potential downsides”. Idemia and Motorola, which provided matches from the authorities’ database, declined to comment.
While the latest research has eased lawmakers’ reservations, the debate continues. The General Services Administration, which oversees federal contractors, said in a report rolled out last month that the primary facial recognition tool was disproportionate to African-Americans in its assessments. The company did not respond to a request for details about the test.
Facial recognition could be reviewed by the U.S. president’s new National Advisory Council on Artificial Intelligence, which last week began assembling a subgroup to understand its use in policing.
America’s first case
Virginia authorized its ban through a process that restricts access to facial recognition builders. This year, corporate lobbyists are poised to advance laws that better balance the liberties of certain individuals with the needs of police investigations, said state Sen. Scott Surovell.
Beginning July 1, police can use facial-recognition instruments that achieve 98 percent or better accuracy on no less than one NIST inspection, with minimal demographic differences.
NIST declined to comment, citing compliance with discussion laws.
Tech critics say the convention is well-intentioned, but not perfect, and must require authorization for the use of facial recognition.
Os Keyes, an Ada Lovelace researcher at the University of Washington, said: “Addressing discriminatory policing by double-checking algorithms is a bit like trying to address police brutality by checking whether guns are racist: strictly speaking, it’s higher than choice, But the real downside is the specific person holding the firearm.”
Virginia bans real-time surveillance, and facial matching cannot be used as a possible trigger for a warrant. Misuse can lead to misdemeanor offenses.
Lobbyist Parker called the legislation “the first in the nation to require the accuracy of facial recognition technology used by law enforcement to be assessed by the U.S. government” and “the most rigorous use algorithm in the nation.”
Former Virginia Representative Lashrecse Aird, who spearheaded last year’s legislation, mentioned that this year’s companies need a model to beat thebeautifulban. “They thought it would ensure a higher sense of responsibility – it’s an improvement, but I don’t know,” she said.
It contrasts sharply with legislation in Washington state that requires businesses to conduct their own assessments beforehand “in operational situations.”
“A Moment of Crisis”
California banned police in 2019 from utilizing facial recognition on unit devices such as body cameras. But that ban expired on Jan. 1, thanks to a provision added by state senators.
Jennifer Jones, a workers attorney with the ACLU of Northern California, noted that now, the growing information experience about retail theft and looting has caught the attention of lawmakers. As a result, the ACLU encountered resistance from law enforcement to make the ban permanent.
“Police departments are harnessing the fear of this crime to build more energy,” Jones said. “This has been happening for many years, and we’re seeing new applied science being driven in times of disaster.”
Activists in New York City are also pressing for a ban on facial recognition despite rising crime rates. Eric Adams, who became mayor in January, mentioned a month later that it was safe to use under current guidelines, while former New York Mayor Bill de Blasio said additional caution was needed.
In West Lafayette, officials twice in the past six months have not issued a ban on facial recognition, citing its value in the investigation. “Banning it or dropping its app is a little shortsighted,” said Mayor John Dennis, a former police officer.
David Sanders, the city councillor behind the ban proposal, said concerns about worsening low morale among officers “dominated the response.”
After the loss in Virginia, the civil liberties team escalated in New Orleans. Ten national groups last week recommended lawmakers strengthen rather than repeal their ban, citing the potential for wrongful arrests based largely on flawed identification. Homegrown group Eye on Surveillance said New Orleans “can’t go back.”