The prototype would comb through websites and photos to piece together where a person lives and works, which social accounts they use and who they may be connected to.
Clearview AI has built and tested InquiryIQ, an AI tool that would pick up where the company’s face search leaves off. That search runs on photos scraped from the web and, by Clearview’s count, is used by more than 2,000 US law enforcement agencies. WIRED found InquiryIQ in files that Clearview’s public login page loads for anyone, no account required.
The company says the tool has never been offered to customers; chief executive Amos Kyler adds that no one in law enforcement has tried it. The files show how it was designed, not whether it works.
From details an investigator picks out of a match, and optionally the person’s age, gender and race, InquiryIQ would chase leads across the web on its own, running face recognition on the photos it finds.
The user can choose which AI model runs the research, and one option is a model from xAI, whose Grok chatbot posted antisemitic messages praising Hitler in 2025. Kyler says the choice was there so engineers could compare models in testing.
Before any finding goes into a profile, the interface warns it may be inaccurate and requires the officer to confirm they checked it independently. In a Minnesota case involving federal surveillance of political activists, defense lawyers obtained a report from Clearview’s existing face search whose user had accepted every match, including one the system itself rated less likely. They allege it included photos of a different man entirely, along with his pregnant wife and young daughter.