This startup collects billions of images from the internet to create a face recognition database

Tram Ho

A face recognition startup is being used by hundreds of law enforcement agencies in the US to solve crimes, but the strange thing is that not many people know the software – including the executive community. France is using it every day!

According to The New York Times, the software called Clearview AI is a collaboration between Ton That Hoan, a Vietnamese-Australian Australian software engineer, moved to the US in 2007, and Richard Schwartz, former aide to the former New York City mayor Rudy Giuliani.

The two met at the Manhattan Institute, a conservative policy institute. Kirenaga Partners, a small New York-based private company, is also one of the original investors in Clearview AI. And Clearview AI itself is also receiving funding from venture capitalist Peter Thiel – the first major investor of Facebook, currently still in the board of directors of the social networking company.

Startup này thu thập hàng tỷ hình ảnh từ internet để tạo cơ sở dữ liệu nhận dạng khuôn mặt - Ảnh 1.

Ton That Hoan – co-founder of Clearview AI

When a user uploads an image to the application, which is being used by more than 600 law enforcement agencies, Clearview AI will scan for matching images from the catalog containing the billions of photos it collects from social networking websites. – of course, this is a violation of the terms of service of photo-collecting websites. It will then display the results to anyone who is performing the search.

Clearview does not disclose which law enforcement agencies are using its tools. In addition to hundreds of law enforcement agencies, the company also licenses software to other private companies. But according to The Times investigative article, both national and local law enforcement agencies have confirmed they are using the software to help solve crimes, from shoplifting in stores. , to murder. Law enforcement agencies say they hardly know who developed Clearview AI, or how the software works!

Facebook said it was investigating Clearview AI after The Times article.

The Times revealed that the founders of the app sell services to law enforcement for as low as $ 2,000. Founders rely on current and retired Republican officials to reach law enforcement, persuade them to use the service at a discount, or in some cases to try it out. free.

Upon analysis, it was discovered that the application’s code was designed to work with augmented reality technology, meaning that someone wearing special glasses would be able to use Clearview AI to immediately. Instantaneous identification of the details in front of you, including the identity and address of a person. Ton That Hoan says his company develops augmented reality technology just for testing and has no plans to launch it to the market.

In a statement, a Facebook spokesperson said the company is investigating Clearview AI.

Collecting Facebook information or listing Facebook information in a category is prohibited by our policies, so we are evaluating claims about the company and will take reasonable action if it is found. see they are violating the rules “.

The spokesperson did not comment on Thiel’s investment in the startup, instead he quoted a statement from Thiel’s spokesperson:

In 2017, Peter gave a talented young founder $ 200,000, which two years later turned into capital for Clearview AI, ” said Jeremiah Hall, Thiel’s spokesperson. ” It was Peter’s only contribution; he didn’t join the company.

Thiel was the sponsor of the Hulk Hogan lawsuit that caused Gawker to go bankrupt. Thiel and Ton That Hoan both received negative reviews regarding Gawker.

Startup này thu thập hàng tỷ hình ảnh từ internet để tạo cơ sở dữ liệu nhận dạng khuôn mặt - Ảnh 2.

Peter Thiel, investor of Clearview AI

Spokespeople for other social media platforms used by Clearview AI, such as Twitter, YouTube, Instagram, and Venmo, confirm that collecting images on these websites is a violation of company policy. Twitter further shared that it was acting against the policy of using images from the platform for facial recognition.

When The Times asked an executive to run a photo of a reporter in the software, representatives of Clearview AI – those who had ignored the reporter’s offer to provide information – contacted the reporters. law enforcement agencies to find out if they’re talking to the media.

This led The Times to conclude that the company was tracking journalist Kashmir Hill, and Clearview AI could see what law enforcement forces were looking for, and when they had been searching before.

Concerns about facial recognition technology have long been at the center of many security issues and potential allegations of racism in technology. In December 2019, a study conducted by the National Institute of Standards and Technology found that facial recognition technology is racist, mainly because it identifies non-white people. and women are inferior to the rest.

On Wednesday, Rep. Alexandria Ocasio-Cortez, a Democrat from New York, expressed fears of facial recognition technology at a meeting of the Senate reform and supervisory committee.

We are seeing things like the Black Mirror in real life,” Ocasio-Cortez said, referring to the British blockbuster series on Netflix with content revolving around crazy aspects of the technology.

Startup này thu thập hàng tỷ hình ảnh từ internet để tạo cơ sở dữ liệu nhận dạng khuôn mặt - Ảnh 3.

Congressman Alexandria Ocasio-Cortez

Despite concerns regarding facial recognition technology, law enforcement agencies across the US are believed to have embraced this controversial technology, but they do not often use it publicly due to the nature of investigation process.

But besides the common risks and controversy with the use of facial recognition technology, Clearview AI also presents new risks, when sensitive images are likely to be shared by law enforcement. software without knowing how the company will handle its data.

The Times said police agencies have been using facial recognition technology for decades, but tools like Clearview AI allow searching in addition to government databases. has long been a restriction on law enforcement’s facial recognition software.

Clearview AI says its software has an efficiency rate of about 75%, but The Times says it is impossible to quantify the number of false positives it provides because it has not been tested by any third party. daddy.

Reference: BusinessInsider

Share the news now

Source : Genk