Clearview AI is testing InquiryIQ, an AI tool for automated digital profiling via facial recognition, aimed at enhancing data collection efficiency.
Clearview AI is in the testing phase of a new artificial intelligence tool, InquiryIQ, aimed at automating the process of gathering personal information through facial recognition. This tool can reportedly aggregate data from the internet, including social media profiles and employment history, as revealed by WIRED.
Journalists uncovered details about InquiryIQ in files sent from Clearview AI's login page, which included code and feature descriptions. However, access to the server-side of InquiryIQ was not granted, preventing independent verification of its functionalities.
Streamlining Data Collection
Traditionally, Clearview AI has used facial recognition to generate leads for investigations, requiring manual verification of any associated data. InquiryIQ aims to enhance this by automating the search for additional information. Once it receives initial data, the system autonomously scours the internet for relevant images and information.
The tool produces a Candidate Graph that outlines potential identities and connections, which may include various personal details such as addresses, phone numbers, and social media accounts. The interface allows users to specify demographic parameters like age and race to refine search results, although the impact of these characteristics on the AI's decisions is not fully understood.
Clearview AI has stated that InquiryIQ has not been used by law enforcement and is not available for client use in its current state.
Despite Clearview's warnings about the potential inaccuracies of automatically gathered information, legal experts like Boston University law professor Woodrow Hartzog have raised concerns about the adequacy of human oversight in this automated process. He cautioned that personnel might become mere formal validators of the AI's findings.
AI Models and Implications for Surveillance
The InquiryIQ interface includes options for various analysis models, with some sourced from SpaceXAI and Amazon Bedrock, although Amazon has clarified that it is not involved in the tool's development. Experts have indicated that such automation could significantly decrease the time investigators spend on data collection.
However, Hartzog also warned about the risks of automation, noting that it could enable broader surveillance by simplifying the process of linking disparate information, which was previously labor-intensive. Andrew Guthrie Ferguson, a professor at George Washington University, pointed out that if the system retains logs of its processes, it could enhance the auditability of investigations, presenting a potential advantage amid the risks.
Clearview AI claims to have amassed a database of over 70 billion images sourced from public internet data, with more than 2,000 U.S. law enforcement agencies reportedly using its technology. Reports from 2023 indicated that American police had employed Clearview AI's facial recognition capabilities nearly 1 million times. The company's CEO, Hoan Ton-That, noted that the database had grown to an estimated 30 billion images. Previously, the French authorities imposed a €20 million fine on Clearview AI for the unlawful processing of personal data and ordered the company to stop collecting information on its citizens.
