The Federal Trade Commission has demanded testimony from Anthropic executives as part of an investigation into whether the company's artificial intelligence systems could harm consumers. The inquiry covers the developer of Claude and forms part of broader federal scrutiny of AI companies, including OpenAI.
An FTC spokesperson confirmed the Anthropic testimony demands to The Hill. The available reports do not identify which executives have been called to testify or provide a timetable for their appearances. The agency has not confirmed whether it has issued comparable testimony demands to OpenAI's leadership.
The requests follow demands for company documents. Axios reported that the FTC sent Anthropic and OpenAI civil investigative demands spanning roughly 50 pages, seeking information about security incidents, consumer privacy, internal safeguards and marketing. Those requests place both the operation of AI products and the claims companies make about them within the investigation's scope.
Seeking executive testimony gives investigators another way to examine corporate decisions alongside written records. Such testimony can help establish how a company assesses risks, communicates problems internally and decides what protections to introduce. The reported demands do not, by themselves, establish that either company violated the law.
The FTC enforces federal prohibitions on unfair or deceptive commercial practices. That authority applies to AI products as it does to other consumer technologies: companies can face scrutiny over misleading representations about a product's capabilities, the handling of personal information or promises about security. An investigation is distinct from an enforcement action, and the reports do not describe a final agency finding against Anthropic.
The distinction between model capabilities and product safeguards is particularly relevant to AI agents. Unlike a chatbot limited to generating responses, an agent can be given tools to interact with websites, software and data. Depending on its permissions, that can allow it to carry out tasks beyond the conversation itself. Access controls, human approval requirements and records of tool use therefore matter when assessing what a system can do and how errors can be contained.
The regulatory scrutiny also coincides with Anthropic's preparations for a public offering. The company has publicly filed offering documents that discuss risks associated with its technology, including cyberattacks, biological weapons development and government surveillance. Such disclosures identify possible risks for investors; they are not findings that those harms have occurred.
For now, the confirmed development is the FTC's effort to obtain testimony from Anthropic's leadership while examining company records. Which executives will testify, what questions they will face and whether the investigation will lead to enforcement remain unresolved.
Sources: The Hill, Stocktwits, UA.NEWS