Clearview AI: Scraping the Internet to Identify Anyone
Clearview AI operates the largest known facial recognition database in the world. They scraped over 30 billion photos from social media platforms, news sites, and public websites to build a system that can identify most Americans from a single photo.
How it was built
Clearview’s founders systematically scraped photos from:
- Facebook, Instagram, Twitter, LinkedIn
- YouTube thumbnails
- Venmo transaction photos
- News articles and press releases
- Any publicly accessible website with photos
They paired each photo with whatever identifying information was available: names, links to social profiles, locations, and associated metadata.
Who uses it
Over 3,100 law enforcement agencies have used Clearview, including:
- FBI
- ICE/DHS
- Hundreds of local police departments
- Interpol
- Law enforcement in at least 27 countries
Officers upload a photo from a surveillance camera, witness phone, or social media, and Clearview returns potential matches with links to the source profiles.
Legal challenges
Clearview has faced lawsuits and regulatory action globally:
- Illinois: Sued under BIPA (Biometric Information Privacy Act). Settled for $40 million
- Australia: Found to have violated the Privacy Act
- UK: Fined 7.5 million pounds by the ICO
- Italy: Fined 20 million euros by the DPA
- France: Fined 20 million euros by CNIL
- Canada: Found to have violated privacy law, ordered to delete Canadian data
Despite these actions, Clearview continues operating. The database exists. The photos aren’t being deleted.
Why this changes the camera calculus
Before Clearview, a surveillance camera recorded anonymous video. After Clearview, any photo from any camera can be fed into a system that returns a name, social profiles, and associated information.
Every camera mapped by Sentinel is potentially a Clearview input. The cameras themselves haven’t changed. But what can be done with their footage has changed fundamentally.
See facial recognition deployments and camera density on Sentinel.