Just a few months ago, Clearview AI faced outrage for scraping photos off social media sites like Facebook to create a near-universal facial recognition system that has been embraced by law enforcement agencies.
Now, the company is again in the spotlight, this time after its CEO, Hoan Ton-That, said it’s in discussions with federal and state agencies to help with contact tracing of people infected with the coronavirus.
Ton-That said federal authorities and three states have communicated with him about using his company’s technology as part of their coronavirus mitigation efforts. NBC News reached out to the governors of all 50 states and several federal agencies, but none have confirmed whether such conversations are taking place, and Ton-That declined to describe the discussions.
“We mainly serve government and law enforcement, and we’ve seen more demand for solutions around coronavirus and how to help things like tracing people who may have had the virus,” Ton-That said. “What we understand now is we’re in the stage where if we are to open up the economy in a way that’s safe for everybody, that we need to be able to test quickly and also trace the people who have been infected and find out who they’ve been in contact with.”
Clearview AI’s contact tracing ambitions, first reported by The Wall Street Journal and confirmed by the CEO in an interview with NBC News on Monday, have already drawn a response from Sen. Ed Markey, D-Mass., who sent a letter Thursday demanding that the company reveal the state and federal agencies it has spoken with about the possibility, as well as the nature and methodology of any role the company might play in a contact tracing program.
“Clearview has failed to demonstrate that it can be trusted to protect Americans’ privacy,” Markey, a member of the Senate Commerce, Science and Transportation Committee, said in a statement. “I’m concerned that if this company becomes involved in our nation’s response to the coronavirus pandemic, its invasive technology will become normalized, and that could spell the end of our ability to move anonymously and freely in public.”
Ton-That said in an emailed statement: “We just received the letter from Senator Markey, for whom we have great respect. We will be responding to him directly.”
Clearview AI relies on images and data taken from social media accounts without users’ explicit permission to put names to even low-resolution photos. The company operated in relative secrecy for a year and has since signed up more than 2,200 government and law enforcement clients around the world, Ton-That said.
The secrecy ended in January, when The New York Times published a profile of the company and detailed its questionable data practices. Since then, the company has come under attack from Facebook for using its photo database to identify millions of people and from politicians like Markey, who opened an investigation into how foreign governments might use Clearview’s services to “suppress their citizens.”
A leaked client list from Clearview AI obtained by BuzzFeed News included paid contracts with Immigration and Customs Enforcement and the U.S. Attorney’s Office for the Southern District of New York, as well as hundreds of police departments. The Secret Service, the Drug Enforcement Administration and the FBI had trial subscriptions, according to the list. Ton-That has previously declined to discuss the company’s client list. ICE confirmed in an email that it does use Clearview AI but didn’t specify whether its use is related to contact tracing.
The company’s political leanings have also been the subject of scrutiny. Ton-That has been connected with conservative provocateurs and allies of President Donald Trump, making him an outlier in the tech industry.
The company’s investors include Peter Thiel, the venture capitalist and Facebook board member who spoke at the 2016 Republican National Convention, whose investment portfolio includes the technology and surveillance companies Anduril and Palantir, both of whom do work for federal agencies. Palantir is also reported to be working with the Department of Health and Human Services to create a coronavirus data platform.
Using technology to track the spread of the coronavirus is already a reality in some parts of the world, including South Korea, Singapore and China, where, depending on the system, a combination of location information, credit card purchase history and other data streams has created the ability to track residents and people with whom they may have been physically close.
Western countries have also been discussing tracing capabilities but with a heavier emphasis on privacy. Three states — Utah, South Dakota and North Dakota — have announced their own contact tracing apps, which will use GPS data to track each phone’s location. But the vast contact tracing program announced in April by Apple and Google uses Bluetooth proximity data, which the companies say is easier to keep private.
But Clearview’s technology could enable more aggressive contact tracing. Ton-That said the company could identify people in surveillance video in settings such as restaurants or stores where they may have come into contact with people who later tested positive for the coronavirus. Anyone appearing in surveillance video could, in theory, be identified by name using Clearview AI’s technology.
Ton-That said that discussions with state and federal authorities are preliminary but that the existing network of surveillance cameras in public and private establishments already provide Clearview AI what the company needs for facial recognition. Ton-That said it would be up to the agencies that use the technology to get appropriate consent for its use.
“A lot of retail spaces and gyms, they already have cameras,” Ton-That said. “And there is the expectation that you’re in a public area, so there’s not necessarily an expectation of privacy. The cameras are already there in case crime happens. They can now already be repurposed to help track anyone else who has had the virus.”
It’s not yet clear whether a contact tracing system can slow the spread of the coronavirus, but even if it can, critics wonder whether one that incorporates technology like Clearview AI’s might forever compromise a user’s privacy.
“In many places across the world, technology has been mixed with what I would call implements of social control that are coercive or quasi-coercive,” said Glenn Cohen, faculty director of the Petrie-Flom Center for Health Law Policy, Biotechnology & Bioethics at Harvard University. “There are big risks about privacy, about the relationship we have to the state, about access to the things we need to have good lives, like employment. And these are all things that I think need to be sorted out in real time rather than say: ‘Oh, there’s this shiny new toy over here. Let’s go ahead and use it, because what harm can it do?'”
Ton-That said that he believes it’s not up to Clearview to deny state or federal authorities the tools it makes.
“It’s really up to the states to see how they put these programs together,” he said. “Even the tech companies and everyone else is saying, ‘Once it’s over, these things should be kind of gone.’ All we do is provide the identification part of the process, and I think there are very strong laws … that prevent the use of medical data for anything else.”