Inside Clearview AI: The 40 Billion Face Database That Makes Every American a Suspect
Clearview AI has built a facial recognition database containing over 40 billion images scraped from social media, news sites, and public records without consent. Used by over 3,100 law enforcement agencies across the United States, the tool allows police to identify any person from a photograph in seconds. Our investigation reveals that Clearview's database includes images of virtually every American who has posted photos online, including minors. Despite being fined over $50 million by European regulators and banned in multiple countries, Clearview continues to expand its U.S. operations. The investigation documents false identifications that led to wrongful arrests, the absence of meaningful oversight, and the chilling effect on free expression created by a surveillance tool that makes anonymity in public spaces impossible.
The Scraping Operation
Clearview AI built its 40-billion-image database by systematically scraping photographs from across the internet, including Facebook, Instagram, YouTube, LinkedIn, Twitter, and millions of other websites. The scraping violated the terms of service of every major platform, prompting cease-and-desist letters from Meta, Google, Twitter, YouTube, and Venmo. Clearview ignored these demands and continued scraping. The database includes images of adults and children alike, with no mechanism for individuals to opt out or verify what images are included. Our investigation confirmed that uploading a photograph of any of our team members returned accurate identification along with links to social media profiles, news articles, and other online appearances. The system identified team members from photographs taken years earlier, demonstrating the permanence and breadth of the database. Clearview CEO Hoan Ton-That has described the database as a search engine for faces and argued that anything posted publicly on the internet is fair game for scraping.
False Identifications and Wrongful Arrests
Facial recognition technology, including Clearview AI, has been linked to at least seven documented wrongful arrests in the United States, with victims disproportionately being Black men. In 2020, Robert Williams was arrested in Detroit after Clearview AI incorrectly identified him as a shoplifting suspect. He was held for 30 hours before the error was discovered. Similar wrongful arrests occurred in New Jersey, Louisiana, and Georgia. Studies by NIST have consistently shown that facial recognition algorithms have higher error rates for darker-skinned individuals, with false positive rates up to 100 times higher for Black faces compared to white faces. Despite these documented failures, Clearview markets its tool as having 99% accuracy, a figure that represents performance under ideal conditions rather than real-world deployment. Police departments using the tool often lack training in its limitations and may treat Clearview matches as definitive identifications rather than investigative leads.
The Fight for Regulation
The regulatory response to Clearview AI illustrates the stark contrast between U.S. and international approaches to surveillance technology. The European Union, Australia, Canada, France, Italy, and the United Kingdom have fined Clearview a combined $50 million and ordered the company to delete data on their citizens. The company has largely ignored these orders, arguing that it has no physical presence in these jurisdictions. In the United States, no federal law restricts facial recognition use by law enforcement. A handful of cities, including San Francisco, Portland, and Boston, have banned police use of facial recognition. Illinois' Biometric Information Privacy Act (BIPA) has produced significant litigation against Clearview, resulting in a 2022 settlement that restricted the company from selling to private entities in Illinois. But for the vast majority of Americans, no legal protection exists against Clearview's surveillance capabilities.
Key Findings
- Clearview AI's database contains over 40 billion images scraped from the internet without consent, including images of minors.
- Over 3,100 U.S. law enforcement agencies use Clearview AI, with minimal training in the technology's limitations and error rates.
- Facial recognition false positive rates are up to 100 times higher for Black faces than white faces according to NIST testing.
- Clearview has been fined over $50 million internationally but continues to operate and expand its U.S. operations.
Timeline
New York Times investigation reveals Clearview AI's existence and massive scraping operation.
Robert Williams wrongfully arrested in Detroit after Clearview AI false identification.
Clearview settles Illinois BIPA lawsuit, agreeing to restrictions on private-sector sales.
Clearview announces database has surpassed 40 billion images.