Clearview AI, the facial recognition company that claims to have amassed a database of more than 3 billion photos scraped from Facebook, YouTube, and millions of other websites, is scrambling to deal with calls for bans from advocacy groups and legal threats. These troubles come after news reports exposed its questionable data practices and misleading statements about working with law enforcement.
Following stories published in the New York Times and BuzzFeed News, the Manhattan-based startup received cease-and-desist letters from Twitter and the New Jersey attorney general. It was also sued in Illinois in a case seeking class-action status.
Despite its legal woes, Clearview continues to contradict itself, according to documents obtained by BuzzFeed News that are inconsistent with what the company has told the public. In one example, the company, whose code of conduct states that law enforcement should only use its software for criminal investigations, encouraged officers to use it on their friends and family members.
“To have these technologies rolled out by police departments without civilian oversight really raises fundamental questions about democratic accountability,” Albert Fox Cahn, a fellow at New York University and the executive director of the Surveillance Technology Oversight Project, told BuzzFeed News.
In the aftermath of revelations about its technology, Clearview has tried to clean up its image by posting informational webpages, creating a blog, and trotting out surrogates for media interviews, including one in which an investor claimed Clearview was working with “over a thousand independent law enforcement agencies.” Previously, Clearview had stated that the number was around 600.
Clearview has also tried to allay concerns that its technology could be abused or used outside the scope of police investigations. In a code of conduct that the company published on its site earlier this month, it said its users should “only use the Services for law enforcement or security purposes that are authorized by their employer and conducted pursuant to their employment.”
It bolstered that idea with a blog post on Jan. 23, which stated, “While many people have advised us that a public version would be more profitable, we have rejected the idea.”
“Clearview exists to help law enforcement agencies solve the toughest cases, and our technology comes with strict guidelines and safeguards to ensure investigators use it for its intended purpose only,” the post stated.
But in a November email to a police lieutenant in Green Bay, Wisconsin, a company representative encouraged a police officer to use the software on himself and his acquaintances.
“Have you tried taking a selfie with Clearview yet?” the email read. “It’s the best way to quickly see the power of Clearview in real time. Try your friends or family. Or a celebrity like Joe Montana or George Clooney.
“Your Clearview account has unlimited searches. So feel free to run wild with your searches,” the email continued. The city of Green Bay would later agree on a $3,000 license with Clearview.
Hoan Ton-That, the CEO of Clearview, claimed in an email that the company has safeguards on its product.
“As as [sic] safeguard we have an administrative tool for Law Enforcement supervisors and administrators to monitor the searches of a particular department,” Ton-That said. “An administrator can revoke access to an account at any time for any inappropriate use.”
Clearview’s previous correspondence with Green Bay police appeared to contradict what Ton-That told BuzzFeed News. In emails obtained by BuzzFeed News, the company told officers that searches “are always private and never stored in our proprietary database, which is totally separate from the photos you search.”“So feel free to run wild with your searches.”
“It’s certainly inconsistent to, on the one hand, claim that this is a law enforcement tool and that there are safeguards — and then to, on the other hand, recommend it being used on friends and family,” Clare Garvie, a senior associate at the Georgetown Law’s Center on Privacy and Technology, told BuzzFeed News.
Clearview has also previously instructed police to act in direct violation of the company’s code of conduct, which was outlined in a blog post on Monday. The post stated that law enforcement agencies were “required” to receive permission from a supervisor before creating accounts.
But in a September email sent to police in Green Bay, the company said there was an “Invite User” button in the Clearview app that can be used to give any officer access to the software. The email encouraged police officers to invite as many people as possible, noting that Clearview would give them a demo account “immediately.”
“Feel free to refer as many officers and investigators as you want,” the email said. “No limits. The more people searching, the more successes.”
“Rewarding loyal customers”
Despite its claim last week that it “exists to help law enforcement agencies,” Clearview has also been working with entities outside of law enforcement. Ton-That told BuzzFeed News on Jan. 23 that Clearview was working with “a handful of private companies who use it for security purposes.” Marketing emails from late last year obtained by BuzzFeed News via a public records request showed the startup aided a Georgia-based bank in a case involving the cashing of fraudulent checks.
Earlier this year, a company representative was slated to speak at a Las Vegas gambling conference about casinos’ use of facial recognition as a way of “rewarding loyal customers and enforcing necessary bans.” Initially, Jessica Medeiros Garrison, whose title was stated on the conference website as Clearview’s vice president of public affairs, was listed on a panel that included the head of surveillance for Las Vegas’ Cosmopolitan hotel. Later versions of the conference schedule and Garrison’s bio removed all mentions of Clearview AI. It is unclear if she actually appeared on the panel.
A company spokesperson said Garrison is “a valued member of the Clearview team” but declined to answer questions on any possible work with casinos.
Cease and desist
Clearview has also faced legal threats from private and government entities. Last week, Twitter sent the company a cease-and-desist letter, noting that its claim to have collected photos from its site was in violation of the social network’s terms of service.
“This type of use (scraping Twitter for people’s images/likeness) is not allowed,” a company spokesperson told BuzzFeed News. The company, which asked Clearview to cease scraping and delete all data collected from Twitter, pointed BuzzFeed News to a part of its developer policy, which states it does not allow its data to be used for facial recognition.
On Friday, Clearview received a similar note from the New Jersey attorney general, who called on state law enforcement agencies to stop using the software. The letter also told Clearview to stop using clips of New Jersey Attorney General Gurbir Grewal in a promotional video on its site that claimed that a New Jersey police department used the software in a child predator sting late last year.
Clearview declined to provide a list of law enforcement agencies that were on free trials or paid contracts, stating that it was more than 600.
“We do not have to be hidden”
That number is lower than what one of Clearview’s investors bragged about on Saturday. David Scalzo, an early investor in Clearview through his firm, Kirenaga Partners, claimed in an interview with Dilbert creator and podcaster Scott Adams that “over a thousand independent law enforcement agencies” were using the software. The investor went on to contradict the company’s public statement that it would not make its tool available to the public, stating “it is inevitable that this digital information will be out there” and “the best thing we can do is get this technology out to everyone.”
EPIC’s letter came after an Illinois resident sued Clearview in a state district court last Wednesday, alleging the software violated the Illinois Biometric Information Privacy Act by collecting the “identifiers and information” — like facial data gathered from photos accumulated from social media — without permission. Under the law, private companies are not allowed to “collect, capture, purchase,” or receive biometric information about a person without their consent.
The complaint, which also alleged that Clearview violated the constitutional rights of all Americans, asked for class-action recognition on behalf of all US citizens, as well as all Illinois residents whose biometric information was collected. When asked, Ton-That did not comment on the lawsuit.
In legal documents given to police, obtained by BuzzFeed News through a public records request, Clearview argued that it was not subject to states’ biometric data laws including those in Illinois. In a memo to the Atlanta Police Department, a lawyer for Clearview argued that because the company’s clients are public agencies, the use of the startup’s technology could not be regulated by state law, which only governs private entities.
Cahn, the executive director of the Surveillance Technology Oversight Project, said that it was “problematic” for Clearview AI to argue it wasn’t beholden to state biometric laws.
“Those laws regulate the commercial use of these sorts of tools, and the idea that somehow this isn’t a commercial application, simply because the customer is the government, makes no sense,” he said. “This is a company with private funders that will be profiting from the use of our information.”
To process a request, however, Clearview is requesting more personal information: “Please submit name, a headshot and a photo of a government-issued ID to facilitate the processing of your request.“ The company declined to say how it would use that information.