The U.S. Postal Service is using facial recognition from controversial company Clearview AI and fake identities online to snoop on Americans, a new report reveals.
Under its law enforcement arm, analysts are reported to use intelligence tools to track social media posts and share their results with other agencies.
"The USPIS said the tools used are 'standard law enforcement techniques'. They say their use is 'strictly controlled relative to the investigation of criminal suspects and criminal activities'."
The Internet Covert Operations Program (iCOP) was first revealed in April when it came under scrutiny for tracking Americans' social media posts ahead of protests. Now it emerges iCOP - which is believed to date back to 2018 - is more far reaching than previously thought.
"Clearview AI CEO Hoan Ton-That has said his company collects only publicly available photos from the open internet that are accessible 'from any computer anywhere in the world.' He said its database cannot be used for surveillance."
At least seven states and nearly two dozen cities have limited government use of Clearview's technology amid fears over civil rights violations, racial bias, and invasion of privacy.
"Rachel Levinson-Waldman, deputy director of the Liberty & National Security Program of the Brennan Center for Justice, said: 'The U.S. Postal Inspection Service appears to be putting significant resources into covert monitoring of social media and the creation and use of undercover accounts.'
The USPS is said to use Clearview AI to help identify unknown targets in an investigation or locate additional social media accounts for known individuals. It also uses Zignal Labs - which runs keyword searches on possible threats - and Nfusion - which creates anonymous online accounts - in its tracking.
"This review of publicly available open source information, including news reports and social media, is one piece of a comprehensive security and threat analysis," they added.
A leaked bulletin first reported by Yahoo News details how the US Postal Inspection Service collected data as part of iCOP. Analysts are said to have looked through posts on Facebook, Parler, and Telegram before flagging 'inflammatory' messages to government agencies.
How Does Facial Recognition Technology Work?
Facial recognition software works by matching real-time images to a previous photograph of a person. Each face has approximately 80 unique nodal points across the eyes, nose, cheeks, and mouth which distinguish one person from another.