Clearview AI Is Testing an AI Tool That Would Let Cops Unearth Your Life Online

WIRED reports that Clearview AI has built and tested InquiryIQ, an unreleased “analyst assistant” designed to take details from a face-recognition hit and automatically fan out across the web — opening pages, analyzing images, and assembling possible employers, aliases, associates, and physical traits into a profile.

Clearview says the tool is a prototype, has never been pitched or shipped to customers, and that no law enforcement user has used it. The company frames it as automating searches detectives already run. The stakes are still real: compressing days of open-source investigation into minutes lowers the cost of fishing expeditions, and generative models can pursue different leads from the same inputs in ways that are hard to reconstruct later.

We’re not surprised Clearview is pushing past “face to name.” The company’s database now exceeds seventy billion images, per its own account, and more than two thousand U.S. law enforcement agencies already use the tech. InquiryIQ is the logical next product move: once you have an identity, automate the biography.

What troubles us is the friction story. Privacy scholars have long argued that investigative labor was itself a check — you couldn’t cheaply rummage everyone’s digital leftovers. Tools like this erase that practical barrier. George Washington’s Andrew Guthrie Ferguson calls it “digital rummaging.” Boston University’s Woodrow Hartzog warns that human-in-the-loop review can degrade into a rubber stamp. A Minnesota case WIRED cites — unrelated to InquiryIQ — already shows Clearview outputs stamped “Accepted” even when labeled less likely, including alleged matches that swept in the wrong man’s family photos.

We’re also watching the model choice. Code WIRED reviewed lists SpaceXAI’s Grok lineage and Amazon Bedrock among options. Clearview’s CEO Amos Kyler says those selectors were for engineers comparing models, not for cops to pick a brain. Even so, testing a chatbot known for inflammatory and unreliable outputs as an investigative informant is a red flag. Defense lawyers will ask the obvious question: would a hallucination-prone system ever be trusted as the basis for probable cause in any other setting?

Useful AI in policing can mean faster leads on real crimes. That benefit is real when evidence is grounded and auditable. The downside if InquiryIQ-style systems become normal is mass enrichment of anyone who ends up in a face search — including people who were never serious suspects — with demographic fields that invite the model to “make smarter decisions” using age, gender, and race.

Our read: treat the “prototype only” line as temporary. Modern AI makes sophisticated mock products cheap to build. The trajectory to watch is whether Clearview ships a governed, logged, independently verified version — or whether another quiet file drop reveals a product already in the wild.

WIRED found InquiryIQ the same way it recently found Flock Safety’s OS Investigate mock-up: in files Clearview’s login page sends to any visitor’s browser. Interface text and warnings do not prove server-side behavior or real-world accuracy.

Clearview’s history includes scraping billions of photos, lawsuits, and founder Hoan Ton-That’s departure; Kyler became CEO in late 2025 and emphasizes controls and auditing.