The US Department of Homeland Security and its law enforcement agencies – Immigration and Customs Enforcement, and Customs and Border Protection – have been sued for failing to respond to Freedom of Information Act requests about their use of Clearview AI’s facial-recognition technology.
Four non-profit organizations focused on civil and immigration rights – the ACLU of Northern California, Mijente, Just Futures Law, and the Immigrant Defense Project – filed the lawsuit [PDF] on Tuesday in northern Cali. They all put in freedom-of-information requests to Uncle Sam over its use of Clearview’s tech last year, and have had no response.
“Facial recognition algorithms erroneously match and misidentify Black and Brown people at a disproportionate rate,” said Jennifer Jones, Technology & Civil Liberties Fellow at the ACLU of Northern California in a statement.
“This is particularly alarming information when coupled with the knowledge that Clearview AI has been added to the arsenal of tools federal immigration enforcement agencies have at their disposal. There is a blatant lack of transparency regarding the extent to which ICE and other agencies use face prints and how they share this information.”
And a spokeswoman representing Just Futures Law told The Register last night:
ICE to see you: Homeland Security’s immigration cops tap up Clearview AI to probe child exploitation, cyber-crime
Facial recognition is a controversial technology. Research has shown that it consistently struggles to correctly identify women and people with darker skin from photographs. Despite this, police departments and law enforcement agencies continue to deploy these technologies, which are built by software companies like Clearview AI.
The New York-based upstart is not off to a good start. It claimed it had amassed a large data set to train its facial-recognition software by scraping more than three billion images of people from their public social network profiles. The idea is that its algorithms can take a photograph previously unseen by the software, and will try to match it with ones stored in its database. If there is a match, the application pulls up the person’s online profiles and details for the cops. The ACLU sued Clearview last year for harvesting images from people’s profiles.
“Clearview AI portends a grim future for us: police officers, ICE agents, and border patrol with the ability to know our identity, our jobs, our relationships just by pointing their phones and taking a photo of our faces,” Jacinta Gonzalez, senior campaign organizer with Mijente, said in a statement [PDF]. “This technology is a potential surveillance nightmare, and the bare minimum in public oversight requires that we have access to Clearview’s government contracts and know how our public agencies are using its tools.
All four non-profit organisations spearheading the lawsuit said they all independently filed FOIA requests to Homeland Security, ICE, and CBP in October. Under America’s freedom-of-information laws, government agencies are required to disclose documents related to specific queries within a reasonable time period, if possible.
After being essentially stonewalled, though, the rights warriors have decided to sue, and have claimed the US government has broken the law by wrongfully withholding agency records. “Plaintiffs have exhausted all applicable administrative remedies,” the lawsuit stated.
The info requests hoped to reveal “past and present use of Clearview AI facial recognition technology, including contracts, policies, procedures for verifying results and auditing use of the technology, and related correspondence between the agencies and Clearview AI.”
“As a matter of policy, US Customs and Border Protection (CBP) does not comment on pending litigation,” a CBP spokesperson told The Register. “Lack of comment should not be construed as agreement or stipulation with any of the allegations.”
ICE told The Register the same. Representatives at Homeland Security were not immediately available for comment. ®