Clearview Is Not Backing Down On Facial Recognition, They’re Doubling Down

When 10 Billion Images Isn’t Enough

Clearview AI is a facial recognition company which quietly collected billions of images from social media and other sites on the web, to develop a product to allow you to instantly find all the images a certain face appears in.  They offered a service which would allow you to submit a picture including a persons face and Clearview AI would send every picture they were in which was posted to the internet and scraped by their systems. 

The system was quietly used by law enforcement agencies to build up a profile of people of interest, but also by private companies and individuals willing to pay for the service.  This changed in 2020 when the New York Times published a report on Clearview AI and their practice of collecting images without permission.  At that time the company claimed over three billion images in their database, which prompted Twitter, Facebook, Google and other companies to send cease and desist orders, which Clearview ignored stating that their were protected under the first amendment regardless of the lack of governmental intervention at that time,

It only got worse from there for privacy advocates as there were several data breaches in 2020, with Clearview’s databases being accessed by attackers and likely shared widely.  There were also numerous court cases in the US launched to prevent various police agencies from using the facial recognition software.  Among the reasons they cited was that Clearview AI’s claim of 98.6% accuracy, based off of methodology that ACLU used to test Amazon Rekognition was wrong, as the ACLU reviewed the tests used and found them to be nothing like their actual tests.

Today there have been several updates, not least of which that their database actually consists of over 10 billion images and they are now searchable by an AI to provide instant results.  In their demonstration they took a photo of a reporter and instantly had results from the past decade of online images in which the reporter appeared.  The history of secretive behaviour and outright lying by Clearview AI, their founders, the police departments and other customers of the company only serve to make these new developments even more worrisome.