Clearview AI develops tool to automate police web investigations using AI search

Clearview AI, the surveillance firm that sparked outrage by harvesting billions of photos from the internet, is quietly building a new tool that would let law enforcement automatically scour the web for information about suspects. The experimental system, called InquiryIQ, would take an investigator's initial findings and systematically fan out across the web—searching pages, analyzing images, and assembling profiles of a person's potential employers, known associates, aliases, and whereabouts.

The tool represents a significant expansion beyond Clearview's core face-recognition service, which police have relied on since 2020 to match faces to identities. Now the company is automating the follow-up detective work that investigators traditionally had to do by hand.

How InquiryIQ Would Work#

According to code reviewed by WIRED, InquiryIQ accepts basic details from an investigator—name, age, gender, race, hair and eye color—and uses those to conduct what Clearview calls "automated discovery and enrichment" of personal data from web sources. The system would run web and image searches, browse web pages, and apply facial recognition to photographs it encounters. As it gathers information, it builds a "Candidate Graph" of possible identities and connections, filling out suspects' profiles with potential addresses, phone numbers, employers, social media accounts, and criminal history.

In its current testing phase, the tool includes an interface option to select different AI models—including one from SpaceXAI's Grok—though Clearview CEO Amos Kyler says this was only for internal comparison during development, not for police to choose which model conducts searches.

The Speed-Efficiency Tradeoff#

Experts say InquiryIQ could compress days or weeks of painstaking detective work into minutes. But that efficiency comes with a dark side: it dramatically lowers the barrier to investigating anyone, making it practical for police to scrutinize people they might never have targeted before. What once required human labor—and thus served as a practical check on surveillance—now happens at machine speed and scale.

"The privacy protections we have were mainly built in a world that assumed a certain amount of friction in the ability of governments to collect information about people," says Boston University privacy scholar Woodrow Hartzog. "There are a lot of rules we never had just because we never needed them."

Law professor Andrew Guthrie Ferguson calls this process "digital rummaging"—assembling scattered digital clues from across the web into a coherent profile of someone's life. While tools like InquiryIQ are not Clearview's invention (companies like ShadowDragon and Fivecast offer similar services), the scale and speed of automation raises new concerns.