Clearview AI Facial Recognition Quietly Tested Grok-Powered Profiling Tool

Clearview AI facial recognition technology has quietly gained a new artificial intelligence layer, and the company wasn’t ready for anyone outside its walls to know yet. According to a Wired investigation, the surveillance firm built and tested a prototype called InquiryIQ that pairs its face-matching system with a model from Elon Musk’s xAI, the company behind the Grok chatbot. The tool has never been publicly announced. It surfaced only after reporters found code buried in files that Clearview’s login page sends to any visitor’s browser before they even sign in.

Key takeaways

• Clearview AI has been testing InquiryIQ, a prototype that uses a model from xAI’s Grok to enrich facial recognition results with associates, social media accounts, and other personal details.
• Clearview’s database now holds between 60 and 70 billion facial images scraped from public web content, with an algorithm claiming over 99% accuracy across demographics per National Institute of Standards and Technology testing.
• Clearview’s customer base is limited to vetted US government and law enforcement agencies under a 2022 settlement with the ACLU.
• In February 2026, Clearview signed a one-year, $225,000 contract with US Customs and Border Protection for tactical targeting and counter-network analysis.
• There is no public disclosure or regulatory filing describing InquiryIQ or its integration with xAI’s Grok model.

Clearview AI introduces InquiryIQ with xAI’s Grok model

InquiryIQ is best understood as an attempt to automate the legwork that investigators normally do by hand after a facial match comes back. Once a Clearview AI facial recognition search returns a name, the new tool is designed to keep working, crawling the open web to build out a fuller picture of that person’s life.

Generative AI integration enhances facial recognition

The system’s interface reportedly tells investigators that adding age, gender, and race can help it make “smarter decisions” while it searches, based on code reviewed by Wired. One of the models Clearview tested for that decision-making came from xAI. Clearview’s chief executive, Amos Kyler, has said the option to select xAI or other model providers was included so engineers could compare performance during internal testing, not so officers could choose which model ran an investigation. “No law enforcement user has ever used it, period,” Kyler said. According to Clearview’s statement to reporters, InquiryIQ remains merely a prototype, one that has never been offered or delivered to any customer, and the company has no current plans to launch it in its existing form.

Functionality includes surfacing associates and social media profiles

Rather than simply returning a match and a name, InquiryIQ is designed to surface contextual intelligence around that identification. The tool can run web and image searches, browse pages, and apply facial recognition to photos it encounters along the way, assembling what it calls a “Candidate Graph” of possible identities, associates, addresses, phone numbers, employers, and aliases. Analysts describe this as a meaningful escalation in what Clearview can offer clients, since it moves the company beyond pure identification and into automated, ongoing profile-building drawn from open-source data.

Data scale, accuracy, and customer restrictions

Clearview’s ability to run a tool like InquiryIQ rests on a facial image library that has grown enormously since the company first drew public attention, alongside accuracy claims backed by federal testing and a customer list narrowed by legal settlement.

Extensive facial image database and NIST accuracy claims

The company’s database now contains between 60 and 70 billion facial images, all scraped from publicly accessible corners of the internet. According to Clearview, testing conducted by the National Institute of Standards and Technology found that its core algorithm achieves an accuracy rate exceeding 99% regardless of demographic group. That combination of scale and claimed precision is part of what makes the addition of a generative AI layer notable: InquiryIQ would not just identify a face but attempt to contextualize it, drawing on a dataset already built from billions of scraped images.

Customer base limited to vetted government agencies

Clearview restricts access to vetted government and law enforcement agencies, a limitation written into its 2022 settlement with the ACLU following years of litigation and regulatory scrutiny over its scraping practices. That settlement remains the main legal guardrail shaping who can use Clearview’s tools, including any future version of InquiryIQ, even as the underlying Clearview AI facial recognition platform keeps expanding its reach among police departments and federal agencies.

Government contracts and regulatory transparency

Clearview’s business with federal agencies keeps growing even as the company stays quiet about its newest AI experiment, a gap that raises real questions about how much oversight surrounds tools built on top of its facial-matching platform.

Recent US Customs and Border Protection contract

Back in February 2026, US Customs and Border Protection awarded Clearview a one-year Clearview government contract valued at $225,000, aimed at tactical targeting and counter-network analysis. This agreement demonstrates that federal agencies still rely heavily on Clearview’s core identification technology despite its earlier controversies, even as the company tests adding generative AI onto that same underlying system.

No public disclosures on InquiryIQ and its AI integration

There are no public records, announcements, or regulatory filings describing InquiryIQ or its integration with xAI’s Grok model, which suggests the tool is either in early internal testing or being piloted quietly with a limited number of agencies. That silence matters. Privacy researchers who reviewed similar automated investigation tools argue that the labor once required to dig through someone’s online life served as a practical check on surveillance. Reducing that labor to an AI-driven search removes friction that many existing privacy protections were built around, without any comparable disclosure requirement stepping in to fill the gap.

The lack of transparency also lands at an awkward moment for xAI’s Grok, a model that has drawn scrutiny in the past for producing extremist and inflammatory outputs. Legal experts who work on Fourth Amendment issues have questioned whether a model prone to unpredictable errors should ever underpin decisions that touch on probable cause, even when a human analyst is technically reviewing the output. Clearview maintains that InquiryIQ is meant only to generate leads for investigators to independently verify, not to make determinations on its own.

FAQ

What is InquiryIQ and how does it enhance Clearview AI’s facial recognition?

InquiryIQ is a prototype tool that layers generative AI from xAI’s Grok model on top of Clearview’s platform, surfacing contextual information such as associates, social media accounts, and personal details alongside a facial recognition match.

Who is allowed to use Clearview AI’s tools, including InquiryIQ?

Clearview AI restricts its customer base to vetted US government and law enforcement agencies, as required under a 2022 settlement with the ACLU.

How accurate is Clearview AI’s facial recognition technology?

Clearview claims its facial recognition algorithm achieves over 99% accuracy across demographics, based on testing conducted by the National Institute of Standards and Technology.

Has Clearview publicly disclosed InquiryIQ or detailed its AI integration?

No. There are no public disclosures or regulatory filings describing InquiryIQ or its use of xAI’s Grok model, which points to either early internal testing or a limited pilot with select agencies.

Article produced with the assistance of artificial intelligence and reviewed by the editorial team.