SAN JOSE, California: Facial recognition software company Clearview AI has appropriated the identities of billions of “unsuspecting” people from websites including social media platforms to sell to police, chilling the right to free speech and endangering immigrants and people of color, a new lawsuit claims.

“Clearview also scrapes images of people that were uploaded without their knowledge or consent, including images posted by friends or relatives and even images of people who inadvertently appear in the backgrounds of photographs taken by strangers,” the suit filed Tuesday in Alameda County Superior Court alleged.

“The sheer volume of online photographs Clearview scrapes to capture faceprints for its database makes it a near certainty that anyone whose photographs are posted to publicly accessible portions of the Internet will have been subjected to surreptitious and nonconsensual faceprinting.”

The suit claims Clearview has “illicitly” and “illegally” collected more than three billion photos of “unsuspecting individuals”, giving it a database nearly seven times bigger than the FBI’s.

“Clearview has provided thousands of governments, government agencies, and private entities access to its database, which they can use to identify people with dissident views, monitor their associations, and track their speech,” the suit alleged. “Its mass surveillance technology disproportionately harms immigrants and communities of colour.”

Clearview CEO Hoan Tom-That said in an emailed statement that while other facial recognition software has misidentified people of colour, “an independent study has indicated that Clearview AI has no racial bias.”

Accurate, nonbiased facial recognition technology can reduce the chances of the wrong person being apprehended by authorities, Tom-That said. “It’s much preferable to have law enforcement accurately identify someone, as opposed to looking for a general description, where wrongful detention, apprehension, and arrests are more likely, especially for those in black and brown communities.”

The American Civil Liberties Association has said that use of Clearview’s technology by law enforcement agencies “will end privacy as we know it” and the group disputed the validity of the company’s racial-bias study.

Floyd Abrams, a prominent free-speech lawyer representing Clearview, said in an emailed statement that the firm “complies with all applicable law and its conduct is fully protected by the First Amendment”.

Major Silicon Valley technology firms including Facebook, Google, YouTube, Twitter and LinkedIn have demanded that New York-based Clearview stop scraping images from their platforms. The suit claims Clearview applies algorithms to the photos to create a “faceprint” that is a person’s unique “biometric signature.”

According to the lawsuit, the San Mateo’s Sheriff’s Office has received access to the software on a trial basis, the Antioch Police Department has bought a license to use it, and US Immigration and Customs Enforcement - an admitted Clearview customer - can use the software even in cities including San Francisco, Berkeley, Alameda and Oakland that have banned public-agency use of facial recognition over fears related to privacy, false identifications and racial bias.

Neither the San Mateo Sheriff’s Office nor Antioch police immediately responded to requests for comment. ICE said in an emailed statement that it did not routinely use facial recognition technology for non-criminal immigration enforcement. The agency said it uses Clearview AI’s software primarily for investigating “child exploitation and other cybercrime cases”. Agents “may review open-source information during the course of a criminal investigation to support the agency’s investigative authorities”, ICE said. “This is an established procedure that is consistent with other law enforcement agencies.”

The plaintiffs in the lawsuit are Alameda County activists Steven Renderos, Valeria Rojas, Reyna Maldonado and Lisa Knox, plus community groups Mijente Support Committee and NorCal Resist Fund. A spokesman for the plaintiffs said in an emailed statement that one of them is in the US under the Deferred Action for Childhood Arrivals program, but did not identify which one.

“Plaintiffs have suffered injury to their peace of mind arising from their fear that they will be retaliated against for their constitutionally protected views regarding policing and immigration,” the suit claims. “They fear surveillance of their immigrant and people of color communities, and they fear being targeted for arrest and deportation.

“Each day that Clearview is allowed to continue its illegal activities, Plaintiffs suffer immediate and irreparable injuries, including chilling of their core First Amendment rights of association and to engage in political speech, injuries to their rights to privacy, injuries to their property rights in their own likenesses and biometric information, and injuries to their peace of mind and wellbeing.”

Clearview users can snap a photo, for example at a political rally, upload it to the firm’s database and instantly see other photos of the person on social media platforms and websites, the suit claims. “The websites often describe the person’s address, employment information, political affiliations, religious activities, and familial and social relationships, among other sensitive information,” the suit alleges. “Clearview’s portable surveillance technology thus provides instantaneous access to almost every aspect of our digital lives.”

The suit cites news reports in claiming that as of February last year, people associated with more than 2,200 companies, law enforcement agencies and other institutions had performed almost a half-million Clearview searches. Clearview said late last year that more than 2,400 police agencies were using its software, the suit noted.

 - TNS