Pressure mounts on DHS to stop using Clearview AI facial recognition

Pressure is mounting on the Department of Homeland Security (DHS) to discontinue — or at least clarify — its relationship with Clearview AI, a controversial facial recognition company best known for scraping billions of images of people from social media sites.

A coalition of nearly 70 immigrant rights, civil liberties and privacy groups called on the agency to “immediately stop” using the company’s technology in a letter sent to Homeland Security Secretary Alejandro Mayorkas on Monday that was first shared with The Hill.

Clearview AI provides software that can identify and provide information about individuals using images of their faces. The technology is marketed to law enforcement but has also been used by private organizations like the NBA

The company — which has received widespread criticism and cease-and-desist letters from most major social media platforms — does not disclose what entities it provides services to, but federal contracts and reporting shows that DHS, Immigrations and Customs Enforcement (ICE) and Customs and Border Patrol (CBP) have all paid for access. 

Immigrant advocacy groups Mijente, Just Futures Law and the Immigrant Defense Project along with the American Civil Liberties Union of Northern California, all of which signed on to Monday’s letter, filed public records requests to the agencies for information about if and how they use Clearview AI’s technology. After not receiving responses, the four organizations filed a lawsuit last week to compel the agencies to turn over any relevant documents.

The groups are concerned that immigration authorities could be abusing the facial recognition technology to locate, arrest and even deport individuals using data that they did not consent to share.

“Having Clearview technology puts so much power into the hands of ICE agents that have demonstrated time and time again how often they go rogue and how hard it is to keep them accountable,” Jacinta Gonzalez, Mijente’s field director, told The Hill.

The groups are also concerned that the tool could be used to punish critics of the immigration system. 

“Imagine what ICE agents could do with this type of technology at a protest or at a demonstration,” Gonzalez explained. “The chilling effect that that can have on how people participate in First Amendment protected activity, how people continue to protest state violence, protest immigration policies … it really has huge implications.”

Immigration agencies have said that they only use the technology for narrow purposes, like targeting perpetrators of child abuse.

A spokesperson for ICE told The Hill that “in general, U.S. Immigration and Customs Enforcement’s (ICE) use of Clearview AI’s facial recognition technology is primarily used by Homeland Security Investigations (HSI)” for “criminal investigations.”

Clearview AI CEO Hoan Ton-That said in a statement that HSI specifically uses the tech for their Child Exploitation Unit. 

The ICE spokesperson told The Hill they could not guarantee the tech had not been used for other purposes by the agency but stressed that doing so would not fall under “normal protocols.”

ICE’s Enforcement and Removal Operations team, which operates inside the U.S., has run between 51 and 100 searches on Clearview AI, according to BuzzFeed News reporting. 

A spokesperson for CBP referred questions about the letter to DHS, which did not immediately return a request for comment.

Julie Mao, director of Just Futures Law, urged skepticism about agency claims that the technology is being used narrowly.

“We really can’t believe the words and statements when it comes to how they actually conduct their surveillance,” she told The Hill. Immigrants rights groups have criticized ICE agents for allegedly impersonating police officers or even normal citizens while conducting raids, for example.

Monday’s letter warns that beyond giving immigration authorities excessive surveillance power, Clearview’s technology could lead to wrongful detentions through inaccurate matches.

The National Institute of Standards and Technology, a federal agency within the Commerce Department, released an expansive study of voluntarily submitted facial recognition systems in 2019 that found the majority of them have “demographic differentials” that can worsen their accuracy based on a person’s age, gender or race. Clearview has not participated in any of those tests.

Clearview says its product is more accurate than competitors in part because of the size of its database, which it claims contains more than 3 billion images. 

The company has provided its services to law enforcement officials around the country, but little is known about whether the technology has worked for identifying individuals. Three cases have been made public in the last year in which Black men were wrongfully arrested based on an inaccurate facial recognition match, although none of the three involved Clearview. 

Monday’s letter warns that there might be biases in Clearview’s tech because of its founders reported ties to white nationalist organizations. 

The groups are optimistic that the Biden administration may be willing to cut off ties with Clearview under pressure, especially since the administration is conducting a full review of DHS’s role in migrant detention. There’s also precedent for terminating any contracts with the company.

Clearview has been deemed illegal by Canada, had its use banned for law enforcement officials in New Jersey and been dropped by the Chicago Police Department. Multiple cities including Portland, Ore.; Boston; San Francisco; and Oakland, Calif., have barred law enforcement from using facial recognition technology more broadly.

“Clearview AI’s continued violation of civil rights and privacy rights provide ample reason to discontinue its use,” the groups wrote.

Updated at 10:44 a.m.