Tomcat
Professional
- Messages
- 2,687
- Reaction score
- 1,036
- Points
- 113
In January, The New York Times published an investigation into Clearview, a revolutionary facial recognition and surveillance application, and in April, HuffPost wrote about the company's ties to the far-right; here are the translations of both articles in abbreviations.
Hoan Ton-That and Clearview AI
In 2017, Mr. Hoan Ton-That - an Australian techie and former model - did something life-changing: He invented a tool that could prevent you from walking the street anonymously, and provided it to hundreds of law enforcement agencies, from local police officers in Florida to the FBI and Department of Homeland Security.
His tiny company, Clearview AI, has developed a revolutionary facial recognition app. You take someone's image, upload it, and you have access to public photos of that person, along with links to where those photos were originally posted. The system, based on a database of more than three billion images that Clearview claims it has collected on Facebook, YouTube, Venmo and a million other sites, goes far beyond anything ever created by the United States government or the giants of Silicon Valley. ...
Federal and state law enforcement agencies said that while they had little understanding of how Clearview worked and who was behind it, they used the app to investigate cases of shoplifting, identity theft, credit card fraud, murder, and sexual exploitation of children. ...
Until now, technology that easily identifies each person by his or her face has been banned due to its potential to dramatically interfere with privacy. Tech companies have refrained from releasing such a tool, even if they were able to create one. In 2011, the chairman of the board of Google said that this was the only technology that the corporation was holding back because it could be used “in the grossly wrong”. Some major cities, such as San Francisco, have banned police from using facial recognition technology.
But without public scrutiny, more than 600 law enforcement agencies began using Clearview last year, according to the company, which declined to provide a list of them.
The computer code behind the app analyzed by The New York Times includes a programming language that associates it with augmented reality glasses. Users could potentially identify every person they saw, from a protest activist to an attractive stranger on the subway, revealing not only their names, but also where they live, what they do, and who they know.
Clearview has not only provided the application to law enforcement, it has also licensed at least a few companies for security purposes.
“The possibilities for this application are endless,” said Eric Goldman, co-director of the High Tech Law School at Santa Clara University. "Imagine a law enforcement bastard stalking potential romantic partners, or a foreign government using an app to dig up secrets about people, blackmail them, or throw them in jail."
Clearview was shrouded in mystery and avoided talking about its limits. When The New York Times began researching the company in November, its website was a blank page showing a nonexistent Manhattan address as the location of the office. One of the company's employees found on LinkedIn, a sales manager named John Goode, turned out to be Mr. Hoan Ton-That using a fake name. For a month, people associated with the company did not respond to the journalist's emails or phone calls.
However, ignoring him, she also followed him. At the request of the journalist, several police officers uploaded his photo to the Clear view application. They were soon called by company representatives and asked if they were in contact with the media, which means that Clearview has the ability and willingness to follow who the law enforcement is looking for.
The Clearview app carries additional risks as law enforcement uploads sensitive photographs to the servers of a company whose ability to protect that data has not been verified.
In February 2017, Indiana State Police began experimenting with Clearview. Using the app, the police solved one case within 20 minutes. The two men got into a fight in the park, and it all ended with one of them shooting the other in the stomach. A bystander recorded the crime on his phone, so the police had a snapshot of the shooter's face, which they fired into Clearview.
The police immediately got the result: the man appeared in a video that someone posted on social media, and his name was included in the caption to the video. "He did not have a driver's license and was not arrested, so he was not found in government databases," said Chuck Cohen, captain of the Indiana State Police at the time.
This man was arrested and charged. Cohen said he probably would not have been identified without the ability to find his face on social media. According to the company, the Indiana State Police became Clearview's first solvent customer. Police declined to comment, saying they tested the Clearview app.
Hoan Ton-That says that the instrument doesn't always work. Most of the photographs in the Clearview database are taken at eye level. Most of the material the police upload comes from surveillance cameras mounted on ceilings or high on walls.
Despite this, as the company said, the tool's ability to find matches reaches 75%. But it is unclear how often it gives false matches because it has not been independently tested.
"We don't have any data to infer how accurate this tool is," said Claire Garvey, a researcher at the Georgetown University Center for Privacy and Technology who has studied the government's use of facial recognition. - The larger the database, the higher the risk of misidentification due to the doppelganger effect. They talk about a massive database of random people they find on the Internet."
But current and former law enforcement officials say the app is effective. “It was important for us to see if it worked or not,” said Cohen, a former Indiana Police Captain.
One of the reasons Clearview is gaining popularity is because of its uniqueness: Facebook and other social networks prohibit collecting user images. Clearview violates the terms of service of the sites.
As the police upload photos of the people they are trying to identify, Clearview has a growing database of faces. The company can also manage the results that the police see. After the organization realized that the journalist had asked the officers to launch his photo through the app, his face was specially tagged by Clearview systems and for a while did not show any matches.
“What they are doing is creepy, but there will be many more such companies. There is no monopoly here, ”said Al Gidari, a professor at Stanford Law School. "Unless there is a very strong federal privacy law, we will all lose."
Hoan Ton-That said Clearview only uses publicly available images. If Facebook changes your privacy settings so that search engines can't find them, your Facebook photos won't be included in the database, he said.
At the same time, the application retains all the images it found earlier, even if they were later deleted, although Hoan Ton-That says the company is working on a tool that will allow people to request deletion of images.
“I don't see how we can take advantage of facial recognition technology without the serious surveillance that comes with it. The only way to stop this is to ban it, ”one of the newspaper's interlocutors concluded.
Clearview AI's link to the far-right
Exclusive HuffPost documents reveal that Hoan Ton-That and several people who worked for the company have deep, long-standing connections with far-right extremists.
Videos and private messages received by HuffPost confirm that former Breitbart writer Charles "Chuck" Johnson and Hoan Ton-That collaborated on far-right extremism in 2016. By early 2017 at the latest, they started working on facial recognition technology. At least two people who worked for Johnson were at Clearview.
Johnson was the centerpiece of the extremist web. And he had many enemies: social media that took away pages from him, news channels that exposed him, a liberal society that he was convinced allowed minorities to enter Harvard at his expense. In total, there were about 400 people on this web, according to McHugh, one of its members, who provided HuffPost with several dozen messages from the group. Its members were something of a racist political saboteur and terraformer with money, as well as a small number of far-right celebrities such as George Zimmerman. The latter joined the channel in June 2016. He sold the weapons that killed Trayvon Martin.
In January 2017, Johnson wrote on Facebook that he was "building algorithms for deportation squads to identify all illegal immigrants." He later bragged to friends and acquaintances that he was working on a powerful facial recognition tool.
It's unclear if Johnson ever spoke to Miller, the "architect" of Trump's brutal immigration policies. But around this time, according to Forbes, Johnson was working behind the scenes with Thiel, a member of the Transition Executive Committee, and recommending far-right candidates for science and technology positions in the new administration. A Politico source confirmed that Johnson attended the transition meetings.
By this point, Johnson had a well-documented track record of bigotry and dirty tricks. In 2015, his Twitter page was permanently deleted after asking for funds to "remove" a black civil rights activist. In 2016, he released an early-right podcast called "Fash the Nation," which detailed the evolutionary traits of Jews and blacks. "They're even dumber," Johnson said of African Americans.
Another Hoan Ton-That guy was Bass. In his youth, he became interested in the hacker culture, and in 2013 he was even arrested on charges of computer harassment in Virginia. The case, however, was closed.
Bass later became a staunch racist. He was looking for work at American Renaissance, a heavyweight white nationalist organization, and belonged to a white nationalist group called DC Helicopters. It was the Washington-based branch of Right Cause, an influential Pro-Trump organization that attracted neo-Nazis. Members of DC helicopter pilots met regularly. One of the branch leaders was a State Department official assigned to the Bureau of Energy Resources who advocated nuclear weapons for the white ethnic group.
Another early salaried employee was Douglas McKee, an unremarkable Middlebury graduate who retired from finance to become an ultra-right "superstar" in 2016 under the pseudonym Ricky Vaughn. The global white supremacist advocate was so successful at spreading pro-Trump, anti-Semitic propaganda and Kremlin disinformation that the MIT Media Lab ranked his Twitter account as one of the top election influencers, ahead of NBC News and the Drudge report.
One early investor was Hal Lambert, a Texas financial manager and major GOP fundraiser close to Ted Cruz who claimed to be on Trump's inaugural committee. Lambert also invested earlier this year in Anduril, a defense contractor that is building autonomous surveillance systems to guard the southern border. Lambert opposed the political left. On the phone, Lambert expressed surprise to hear about extremists associated with Clearview. But he admitted that he knew Johnson and knew Hoan Ton-That before Clearview.
To grow the business, Clearview hired Jessica Medeiros Garrison, former executive director of the Republican Association of Attorneys General. The company also retained Paul Clement, a former US attorney general and permanent right-wing attorney on the Supreme Court.
Clement provided legal cover for police using the app to get around civil rights issues. He wrote that Clearview does not violate the Fourth Amendment of the US Constitution because posting something on social media removes any expectations of privacy. Clement also responded to the expected argument that facial recognition led to issues of racial bias. Most of the algorithms tested in 2019 by the National Institute of Standards and Technology are more likely to falsely identify female and black faces than white male faces. Clement wrote that Clearview is more accurate than existing technology and uses "non-race based algorithms." True, he did not provide any evidence to support his claims.
As Clearview issued licenses to police departments, far-right extremists in the company increasingly interacted with law enforcement. In a December 2019 email received by BuzzFeed, Clearview employee Marko Djukic presented a free trial of this technology at a nationwide police congress.
Clearview allowed investors and Trump-affiliated elites to use an unregulated database app. Clearview has opened an account for the company of former Trump campaign spokesman Jason Miller. According to BuzzFeed, the organization conducted about 20 searches.
Scandal
Following the publication of The New York Times, New Jersey's attorney general ordered all police officers in the state to stop using this tool. Two Democratic senators have introduced a bill to impose a moratorium on the use of facial recognition by government officials and contractors until Congress can regulate the technology.
Social media demanded that Clearview stop collecting photos and data.
Plenty of users have filed lawsuits against Clearview.
Hoan Ton-That and Clearview AI
In 2017, Mr. Hoan Ton-That - an Australian techie and former model - did something life-changing: He invented a tool that could prevent you from walking the street anonymously, and provided it to hundreds of law enforcement agencies, from local police officers in Florida to the FBI and Department of Homeland Security.
His tiny company, Clearview AI, has developed a revolutionary facial recognition app. You take someone's image, upload it, and you have access to public photos of that person, along with links to where those photos were originally posted. The system, based on a database of more than three billion images that Clearview claims it has collected on Facebook, YouTube, Venmo and a million other sites, goes far beyond anything ever created by the United States government or the giants of Silicon Valley. ...
Federal and state law enforcement agencies said that while they had little understanding of how Clearview worked and who was behind it, they used the app to investigate cases of shoplifting, identity theft, credit card fraud, murder, and sexual exploitation of children. ...
Until now, technology that easily identifies each person by his or her face has been banned due to its potential to dramatically interfere with privacy. Tech companies have refrained from releasing such a tool, even if they were able to create one. In 2011, the chairman of the board of Google said that this was the only technology that the corporation was holding back because it could be used “in the grossly wrong”. Some major cities, such as San Francisco, have banned police from using facial recognition technology.
But without public scrutiny, more than 600 law enforcement agencies began using Clearview last year, according to the company, which declined to provide a list of them.
The computer code behind the app analyzed by The New York Times includes a programming language that associates it with augmented reality glasses. Users could potentially identify every person they saw, from a protest activist to an attractive stranger on the subway, revealing not only their names, but also where they live, what they do, and who they know.
Clearview has not only provided the application to law enforcement, it has also licensed at least a few companies for security purposes.
“The possibilities for this application are endless,” said Eric Goldman, co-director of the High Tech Law School at Santa Clara University. "Imagine a law enforcement bastard stalking potential romantic partners, or a foreign government using an app to dig up secrets about people, blackmail them, or throw them in jail."
Clearview was shrouded in mystery and avoided talking about its limits. When The New York Times began researching the company in November, its website was a blank page showing a nonexistent Manhattan address as the location of the office. One of the company's employees found on LinkedIn, a sales manager named John Goode, turned out to be Mr. Hoan Ton-That using a fake name. For a month, people associated with the company did not respond to the journalist's emails or phone calls.
However, ignoring him, she also followed him. At the request of the journalist, several police officers uploaded his photo to the Clear view application. They were soon called by company representatives and asked if they were in contact with the media, which means that Clearview has the ability and willingness to follow who the law enforcement is looking for.
The Clearview app carries additional risks as law enforcement uploads sensitive photographs to the servers of a company whose ability to protect that data has not been verified.
In February 2017, Indiana State Police began experimenting with Clearview. Using the app, the police solved one case within 20 minutes. The two men got into a fight in the park, and it all ended with one of them shooting the other in the stomach. A bystander recorded the crime on his phone, so the police had a snapshot of the shooter's face, which they fired into Clearview.
The police immediately got the result: the man appeared in a video that someone posted on social media, and his name was included in the caption to the video. "He did not have a driver's license and was not arrested, so he was not found in government databases," said Chuck Cohen, captain of the Indiana State Police at the time.
This man was arrested and charged. Cohen said he probably would not have been identified without the ability to find his face on social media. According to the company, the Indiana State Police became Clearview's first solvent customer. Police declined to comment, saying they tested the Clearview app.
Hoan Ton-That says that the instrument doesn't always work. Most of the photographs in the Clearview database are taken at eye level. Most of the material the police upload comes from surveillance cameras mounted on ceilings or high on walls.
Despite this, as the company said, the tool's ability to find matches reaches 75%. But it is unclear how often it gives false matches because it has not been independently tested.
"We don't have any data to infer how accurate this tool is," said Claire Garvey, a researcher at the Georgetown University Center for Privacy and Technology who has studied the government's use of facial recognition. - The larger the database, the higher the risk of misidentification due to the doppelganger effect. They talk about a massive database of random people they find on the Internet."
But current and former law enforcement officials say the app is effective. “It was important for us to see if it worked or not,” said Cohen, a former Indiana Police Captain.
One of the reasons Clearview is gaining popularity is because of its uniqueness: Facebook and other social networks prohibit collecting user images. Clearview violates the terms of service of the sites.
As the police upload photos of the people they are trying to identify, Clearview has a growing database of faces. The company can also manage the results that the police see. After the organization realized that the journalist had asked the officers to launch his photo through the app, his face was specially tagged by Clearview systems and for a while did not show any matches.
“What they are doing is creepy, but there will be many more such companies. There is no monopoly here, ”said Al Gidari, a professor at Stanford Law School. "Unless there is a very strong federal privacy law, we will all lose."
Hoan Ton-That said Clearview only uses publicly available images. If Facebook changes your privacy settings so that search engines can't find them, your Facebook photos won't be included in the database, he said.
At the same time, the application retains all the images it found earlier, even if they were later deleted, although Hoan Ton-That says the company is working on a tool that will allow people to request deletion of images.
“I don't see how we can take advantage of facial recognition technology without the serious surveillance that comes with it. The only way to stop this is to ban it, ”one of the newspaper's interlocutors concluded.
Clearview AI's link to the far-right
Exclusive HuffPost documents reveal that Hoan Ton-That and several people who worked for the company have deep, long-standing connections with far-right extremists.
Videos and private messages received by HuffPost confirm that former Breitbart writer Charles "Chuck" Johnson and Hoan Ton-That collaborated on far-right extremism in 2016. By early 2017 at the latest, they started working on facial recognition technology. At least two people who worked for Johnson were at Clearview.
Johnson was the centerpiece of the extremist web. And he had many enemies: social media that took away pages from him, news channels that exposed him, a liberal society that he was convinced allowed minorities to enter Harvard at his expense. In total, there were about 400 people on this web, according to McHugh, one of its members, who provided HuffPost with several dozen messages from the group. Its members were something of a racist political saboteur and terraformer with money, as well as a small number of far-right celebrities such as George Zimmerman. The latter joined the channel in June 2016. He sold the weapons that killed Trayvon Martin.
In January 2017, Johnson wrote on Facebook that he was "building algorithms for deportation squads to identify all illegal immigrants." He later bragged to friends and acquaintances that he was working on a powerful facial recognition tool.
It's unclear if Johnson ever spoke to Miller, the "architect" of Trump's brutal immigration policies. But around this time, according to Forbes, Johnson was working behind the scenes with Thiel, a member of the Transition Executive Committee, and recommending far-right candidates for science and technology positions in the new administration. A Politico source confirmed that Johnson attended the transition meetings.
By this point, Johnson had a well-documented track record of bigotry and dirty tricks. In 2015, his Twitter page was permanently deleted after asking for funds to "remove" a black civil rights activist. In 2016, he released an early-right podcast called "Fash the Nation," which detailed the evolutionary traits of Jews and blacks. "They're even dumber," Johnson said of African Americans.
Another Hoan Ton-That guy was Bass. In his youth, he became interested in the hacker culture, and in 2013 he was even arrested on charges of computer harassment in Virginia. The case, however, was closed.
Bass later became a staunch racist. He was looking for work at American Renaissance, a heavyweight white nationalist organization, and belonged to a white nationalist group called DC Helicopters. It was the Washington-based branch of Right Cause, an influential Pro-Trump organization that attracted neo-Nazis. Members of DC helicopter pilots met regularly. One of the branch leaders was a State Department official assigned to the Bureau of Energy Resources who advocated nuclear weapons for the white ethnic group.
Another early salaried employee was Douglas McKee, an unremarkable Middlebury graduate who retired from finance to become an ultra-right "superstar" in 2016 under the pseudonym Ricky Vaughn. The global white supremacist advocate was so successful at spreading pro-Trump, anti-Semitic propaganda and Kremlin disinformation that the MIT Media Lab ranked his Twitter account as one of the top election influencers, ahead of NBC News and the Drudge report.
One early investor was Hal Lambert, a Texas financial manager and major GOP fundraiser close to Ted Cruz who claimed to be on Trump's inaugural committee. Lambert also invested earlier this year in Anduril, a defense contractor that is building autonomous surveillance systems to guard the southern border. Lambert opposed the political left. On the phone, Lambert expressed surprise to hear about extremists associated with Clearview. But he admitted that he knew Johnson and knew Hoan Ton-That before Clearview.
To grow the business, Clearview hired Jessica Medeiros Garrison, former executive director of the Republican Association of Attorneys General. The company also retained Paul Clement, a former US attorney general and permanent right-wing attorney on the Supreme Court.
Clement provided legal cover for police using the app to get around civil rights issues. He wrote that Clearview does not violate the Fourth Amendment of the US Constitution because posting something on social media removes any expectations of privacy. Clement also responded to the expected argument that facial recognition led to issues of racial bias. Most of the algorithms tested in 2019 by the National Institute of Standards and Technology are more likely to falsely identify female and black faces than white male faces. Clement wrote that Clearview is more accurate than existing technology and uses "non-race based algorithms." True, he did not provide any evidence to support his claims.
As Clearview issued licenses to police departments, far-right extremists in the company increasingly interacted with law enforcement. In a December 2019 email received by BuzzFeed, Clearview employee Marko Djukic presented a free trial of this technology at a nationwide police congress.
Clearview allowed investors and Trump-affiliated elites to use an unregulated database app. Clearview has opened an account for the company of former Trump campaign spokesman Jason Miller. According to BuzzFeed, the organization conducted about 20 searches.
Scandal
Following the publication of The New York Times, New Jersey's attorney general ordered all police officers in the state to stop using this tool. Two Democratic senators have introduced a bill to impose a moratorium on the use of facial recognition by government officials and contractors until Congress can regulate the technology.
Social media demanded that Clearview stop collecting photos and data.
Plenty of users have filed lawsuits against Clearview.
