Siding against controversial facial-recognition outfit Clearview AI, a Vermont judge has refused to dismiss state Attorney General Thomas Donovan's claims that the company violated a state consumer protection law.
The ruling, issued late last week by Chittenden County Superior Court Judge Helen Toor, allows Donovan to attempt to prove that Clearview engaged in an “unfair” practice by allegedly using facial-recognition technology on photos of Vermont residents.
While Toor is allowing Donovan to proceed with most claims, she narrowed the case by dismissing allegations that Clearview violated a state law dealing with the fraudulent acquisition of data.
Clearview attorney Tor Ekeland stated the company was pleased that Toor dismissed that claim. He added that the company contends “the facts as presented by Vermont are not true.”
The decision comes in a lawsuit filed in March by Donovan, who alleged Clearview's acts “are unethical, immoral, and against all social norms and policies,” and who sought an injunction requiring “that the people of Vermont immediately be set free from Clearview's dystopian surveillance database.”
Clearview, which came to public attention earlier this year, says it offers a facial-recognition database to police departments and other “security professionals.”
The company hasn't publicly disclosed names of its customers, but BuzzFeed reported that Clearview's clients include US Immigration and Customs Enforcement, the US Attorney's Office for the Southern District of New York, and Macy’s.
Clearview reportedly created the database by scraping billions of photos from Twitter, Facebook and other companies, then deploying facial-recognition technology to create a faceprint database.
In April, Clearview urged Toor to dismiss Donovan's claims for several reasons, including that it wasn't on notice that its alleged facial-recognition practices could be considered unlawful under the state's consumer protection law.
Clearview argued it was "obviously unfair to prosecute someone for what a person of ordinary intelligence had no fair notice was illegal," adding that Donovan's claims involved a "novel theory of a long distance, digital invasion of the privacy of unnamed Vermont consumers."
"This court should not permit the Attorney General use its vague policy yearnings to arbitrarily prosecute Clearview AI," the company wrote.
Toor rejected that argument, writing that the state's Consumer Protection Act was broad enough to cover emerging technology.
“It would make no sense for the legislature to have to pass a new law as soon as any new technology is invented and brought to market,” she wrote.
“Clearview AI's algorithmic search engines, including its facial biometric algorithm, are protected speech because they convey information in reaction to search queries from Clearview's users,” the company wrote in papers filed with Toor in April.
Toor said in her ruling that even if Clearview's app is a form of speech, Vermont wouldn't necessarily violate the First Amendment by regulating the company.
“The regulation sought by the state here is content-neutral,” she wrote, adding that content-neutral restrictions are valid as long as they further a “substantial” government interest.