Until recently, Hoan Ton-The biggest hit was an app that enabled people to put Donald Trump's distinctive yellow hair on their own photos.
When Ton-Dat did something memorable: He invented a tool that could end your ability to walk the streets anonymously and provided it to hundreds of law enforcement agencies.
His small company, Clearview AI, came up with a groundbreaking face recognition app. You take a photo of a person, upload it and you will see public photos of that person, along with links to where those photos appeared.
The system – whose backbone consists of a database of more than 3 billion photos that Clearview claims to have scraped from Facebook, YouTube, Venmo and millions of other websites – goes far beyond what the US government or Silicon Valley giants have ever had. is built.
The federal and state law enforcement officials said they had limited knowledge of how Clearview works and who is behind it, but that they had used the app to help resolve shop theft, identity theft, credit card fraud, murder, and child sexual exploitation.
Until now, technology that easily identifies everyone based on his or her face has been taboo because of the radical erosion of privacy.
But without public scrutiny, over 600 law enforcement agencies started using Clearview last year, according to the company, which refused to provide a list. The computer code that underlies his app, analyzed by The New York Times, includes programming language to link it to augmented reality glasses; users could potentially identify every person they saw.
Clearview has also licensed the app to at least a handful of companies for security purposes.
“The weaponry options are endless,” said Eric Goldman, co-director of the High Tech Law Institute at Santa Clara University. “Imagine that a rogue state agent wants to stalk potential romantic partners, or that a foreign government uses this to dig up secrets about people to blackmail them or throw them in jail.”
Clearview is enveloped in mystery and avoids the debate about pioneering technology. When I started investigating the company in November, the website was a bare page with a non-existent Manhattan address as a business location. The one employee of the company who was on LinkedIn, a sales manager named “John Good “, turned out to be Ton – that, using a fake name. For a month, people affiliated with the company would not send my emails or phone calls back.
While the company avoided me, it also kept an eye on me. At my request, a number of police officers had taken my photo through the Clearview app. They quickly received phone calls from company representatives asking if they were talking to the media – a sign that Clearview has the ability and in this case the appetite to monitor who the law enforcement is looking for .
Face recognition technology has always been controversial. The Clearview app carries additional risk because law enforcement agencies upload sensitive photos to the servers of a company whose ability to protect its data has not yet been tested.
The company finally started answering my questions and said its earlier silence was typical of an early stage of stealth mode startup. Ton-Dat acknowledged designing a prototype for use with augmented reality glasses, but said the company had no plans to release it. And he said my photo had rung alarm bells because the app “flags potentially abnormal search behavior” to prevent users from what it deemed “inappropriate searches”.
In addition to Ton-That, Clearview was founded by Richard Schwartz – who was an assistant to Rudy Giuliani when he was mayor of New York – and financially supported by Peter Thiel, a venture capitalist behind Facebook and Palantir.
Another early investor is a small firm called Kirenaga Partners. The founder, David Scalzo, rejected his concern about Clearview, which made the internet searchable on the face, and said it was a valuable tool for solving crimes.
“I have come to the conclusion that because the information is constantly increasing, there will never be privacy,” Scalzo said. “Laws must determine what is legal, but you cannot prohibit technology.”
Addicted to AI
Ton-That, 31, grew up far from Silicon Valley, in his native Australia. In 2007 he stopped studying and moved to San Francisco. The iPhone had just arrived, and his goal was to start what he was expecting a vibrant market for social media apps.
In 2015, he spun Trump Hair, which added Trump's distinctive coif to people in a photo, and photo sharing program. Both were buzzy.
Ton-The move to New York in 2016. He started reading academic papers on artificial intelligence, image recognition and machine learning.
Schwartz and Ton-Dat met in 2016 at a book event at the Manhattan Institute, a conservative think tank. Schwartz, now 61, had amassed an impressive Rolodex who worked for Giuliani in the 1990s. The two soon decided to go on the face recognition bus …
Police departments have had access to face recognition tools for nearly 20 years, but they have historically been limited to searching for government-provided images, such as photos of mugs and driver's licenses.
Ton … that wanted to go much further than that. He started recruiting a few engineers in 2016. One helped design a program that can automatically collect images of people's faces on the internet, such as employment sites and social networks. Representatives of those companies said that their policies prohibit such scraping.
Another engineer was hired to perfect a face recognition algorithm derived from academic papers. The result: a system that uses what Ton-De has described as a “state-of-the-art neural net” to convert all images into mathematical formulas, or vectors based on facial geometry – such as how far the eyes of a person are apart.
Clearview has created a large folder that clusters all photos with similar vectors in “neighborhoods “. When a user uploads a photo of a face into the Clearview system, it converts the face into a vector and then displays all the scraped photos that are stored near that vector – along with links to the sites where those photos came from.
Clearview is still small and, according to Pitchbook, a website that keeps track of investments in startups, has raised 7 million dollars from investors. The company refused to confirm the amount.
Go Viral with law enforcement
In February the Indiana State Police started experimenting with Clearview. They solved a case within 20 minutes of using the app. Two men had a fight in a park, and it ended when one shot the other in the stomach. A bystander recorded the crime on a telephone, so the police had a still life from the shooter to walk through the Clearview app.
They immediately got a match: The man appeared in a video that someone had posted on social media, and his name was recorded in a caption on the video. “He didn't have a driver's license and wasn't arrested as an adult, so he wasn't in government databases,” said Chuck Cohen, an Indiana State Police captain at the time.
“The man was arrested and charged; Cohen said he probably would not have been identified without the ability to look for social media for his face. The Indiana State Police became the first paying customer of Clearview, according to the company. (Police declined to comment beyond saying that they tested Clearview's app.)
The most efficient sales technique of the company offered 30 days free trials to officials. Ton … that finally had its viral blow.
Federal law enforcement, including the FBI and the Department of Homeland Security, are trying it, as are Canadian law enforcement authorities, according to the company and government officials.
Ton, that said the instrument doesn't always work. Most of the photos in Clearview's database are taken at eye level. Much of the material that the police uploads is from surveillance cameras mounted on ceilings or high on walls.
Despite that, the company said, its tool finds similarities up to 75% of the time.
One of the reasons why Clearview informs itself is that its service is unique. That's because Facebook and other social media sites prohibit people from scraping images of users; Clearview violates the service conditions of the sites.
Some law enforcement officials said they did not realize that the photos they uploaded were sent to the Clearview servers and stored there. Clearview tries to resolve the concerns with a FAQ document that is given to potential customers and which states that customer service representatives will not look at the photos the police upload.
Clearview has also hired Paul Clement, an American advocate general under President George W. Bush, to alleviate concerns about the legality of the app.
In an August note that Clearview provided to potential customers, including the Atlanta Police Department and the Pinellas County Sheriff Office in Florida, Clement said law enforcement agencies “did not Violate the constitution or relevant existing state biometric and privacy laws when using Clearview for its intended purpose. ”
Clement, now a partner at Kirkland & Ellis, wrote that the authorities do not have to tell the suspects that they have been identified via Clearview as long as it is not the only basis for getting an arrest warrant for them. Clement did not respond to multiple requests for comment.
The memo seemed to be effective; the Atlanta police and the Pinellas County Sheriff office soon began using Clearview.
Woodrow Hartzog, a professor of law and computer science at Northeastern University in Boston, sees Clearview as the latest proof that face recognition should be banned in the United States.
“We based ourselves on the industry's efforts to control itself and not embrace such high-risk technology, but now these dams break because so much money is on the table,” Hartzog said. “I don't see a future where we will reap the benefits of face recognition technology without the crippling abuse of surveillance that comes with it. The only way to stop itTags: #ArtificialIntelligence, #latestNewsAI, #researchAi