Trouble continues for the Clearview AI start-up, following an investigation by the New York Times on January 18. This company, hitherto discreet, has designed a gigantic database of images of people gleaned from the Internet: Facebook, LinkedIn, Instagram, Twitter, YouTube. It sells software to the American police force, which can, from ‘a simple image of someone, look for equivalences among these three billion images, to identify individuals. Nearly 600 American police services were won over by the proposal.
Facial recognition databases can rely on mugshots or drivers’ licenses for faces or can be open sourced from the streets and cyber highways of the United States. But Clearview AI uses open-source images, and their self-reported three billion-plus image inventory is many times larger than comparable collections and the most extensive set we’ve seen available.
But this one, and the noise that followed the New York Times investigation, did not please Twitter, Facebook, and Google. In the wake of the publication, Twitter demanded that Clearview AI stop using the images found on its platform, arguing that its regulations prohibited this. Google and Facebook joined the other social networks, CBS News learned Wednesday, February 5. The two companies have given notice to Clearview AI not to put an end to any image collection on their sites.
Google, which owns YouTube, said that “YouTube’s policies explicitly prohibit the collection of data that can be used to identify a person. “The company also regrets that Clearview AI compared, to defend itself, its activity with that of the Google search engine. Clearview AI founder Hoan Ton-That, for example, said on CBS: “Google can extract data from all websites. So, if it’s public, if it’s available, it can be in Google’s search engine, and it can be in ours too.”
To which Google responded:
“Most websites want to appear in our search engine, and we give webmasters control over what information on their sites they want to be indexed in our results, they also have the option of nothing being indexed. Clearview AI has secretly collected images of individuals without their consent. “
For its part, Hoan Ton-That A self-taught engineer, co-founder, and CEO of Clearview -Believes that it is acting legally. “Our system designed to take only publicly available information,” he said. Launched in 2017, Clearview AI initially was funded by multi-billionaire Peter Thiel, founder of Palantir, a member of the Facebook board of directors.