Twitter orders Clearview AI to delete and cease scraping images for biometric data

It may be true that, as Internet entrepreneur Hoan Ton-That has said, “a lot of people” are scraping social-media subscribers’ photographs for use in biometric facial recognition applications but Twitter is not pleased about it.

Earlier this week, The New York Times devoted a lengthy story to Ton-That and his startup, Clearview AI, which has copied and re-used — or scraped — 3 billion face images from millions of web sites, including Twitter, Facebook, Instagram and YouTube.

Yesterday, in the aftermath of the article, Twitter informed Clearview AI executives that they are violating Twitter’s terms-of-service policies prohibiting its actions, according to fresh reporting by The Times. A Jan. 22 article states that the social-media giant demanded executives stop scraping site images and “destroy data that it had previously collected” from Twitter.

The images were scraped without their owners’ permission or notice, and placed in a database behind a new facial recognition application from Clearview AI. The application has been licensed by 600 law-enforcement agencies, a small sample of which say it has helped solve crimes in as little as 20 minutes.

There are few external controls on how Ton-That’s company can use the publicly available photos that he considers his to exploit. And if he has internal controls on how Clearview AI or clients use the images, he is not saying.

If the newspaper’s 3 billion photo estimate is accurate, it could be that only the Chinese government has a bigger database of private individuals.

Unsurprisingly, privacy and civil-liberties proponents are quoted in the story questioning the wisdom of Clearview’s business model.

Article Topics

biometrics  |  Clearview AI  |  dataset  |  face photo  |  facial recognition  |  police  |  social media