Clearview AI Tests Tool to Auto-Generate Suspect Profiles From Web

Clearview AI, the controversial facial recognition company that scraped billions of photos from the internet, has developed an experimental AI tool that could fundamentally change how police investigations work. The prototype system, called InquiryIQ, is designed to automatically search the web and assemble detailed profiles of individuals under investigation, according to WIRED, which discovered the tool's interface in publicly accessible code files.

The unreleased system would take information from a Clearview facial recognition search and then autonomously browse websites, analyze images, and compile data about a person's possible employers, physical characteristics, aliases, associates, and addresses. According to interface code analyzed by WIRED, the tool can accept demographic inputs like age, gender, and race to help make "smarter decisions" during its searches.

Why it matters

InquiryIQ represents a shift from identifying faces to automating the broader investigative process. Privacy scholars warn that by eliminating the labor traditionally required to investigate someone, such tools remove practical limits that once constrained surveillance. When detective work that might take days can be compressed into minutes, police can afford to investigate far more people—including those who might never have warranted scrutiny under traditional resource constraints. The technology also raises questions about accountability: generative AI can produce different results from identical starting points, potentially making it difficult to reconstruct why investigators pursued certain leads.

Testing with controversial AI models

During development, Clearview tested multiple large language models to power InquiryIQ's searches, including Grok from SpaceXAI, the Elon Musk company that merged with SpaceX in February. Grok has drawn repeated criticism for producing racist, extremist, and inflammatory content. In 2025, unauthorized changes to Grok's system caused it to inject claims about supposed "white genocide" into unrelated conversations, and later modifications led it to generate antisemitic posts and praise for Adolf Hitler.

Clearview CEO Amos Kyler told WIRED that the model selector visible in the interface was for internal testing only, not for police to choose which AI runs their searches. He emphasized that no law enforcement agency has used InquiryIQ and the company has no current plans to release it in its present form.

Automation and accountability concerns

Legal experts describe tools like InquiryIQ as "digital rummaging" that connects scattered information previously difficult to link. Andrew Guthrie Ferguson, a George Washington University law professor, notes that while such systems could leave clearer audit trails than traditional detective work, history suggests police technology typically gets deployed before governance frameworks are established.

The InquiryIQ interface warns users that automatically generated data "may or may not be accurate" and requires officers to attest they independently verified findings before accepting them. But privacy scholar Woodrow Hartzog argues that human review offers "a little bit of a cold comfort," as investigators may eventually defer to automated systems until human oversight becomes "a sort of rubber stamp."

Clearview, which now claims a database exceeding 70 billion images used by more than 2,000 law enforcement agencies nationwide, says InquiryIQ is merely a prototype for testing investigative leads. Kyler describes the company's current focus as "refinement" and ensuring products meet customer expectations for integrity—a marked shift from the boundary-pushing approach that defined Clearview's early years.

These details were first reported by WIRED, which discovered InquiryIQ by analyzing code files that Clearview's login page sends to visitors' browsers.