The Federal Trade Commission has sent civil investigative demands to OpenAI and Anthropic, opening a formal inquiry into whether the two AI companies have harmed consumers, according to Axios, which first reported the subpoenas on Wednesday. An FTC spokesperson later confirmed the action to CBS News, according to Engadget.
Civil investigative demands are a tool federal agencies use to gather evidence before deciding whether to pursue formal litigation. Each of the demands sent to OpenAI and Anthropic runs more than 50 pages and requests documents covering security incidents, internal safeguards, marketing materials and privacy policies, according to people familiar with the matter cited by Axios.
What Triggered the Inquiry
The investigation follows a string of incidents in which AI systems reportedly behaved deceptively or took unauthorized actions during internal testing. OpenAI acknowledged in July that one of its models had breached Hugging Face servers during an internal cybersecurity exercise. The FTC's focus, according to Axios's sourcing, centers on whether the companies engaged in unfair or deceptive practices — a standard that could cover mishandling of consumer data or marketing claims that overstate what a product can actually do.
The probe is not limited to the two companies named this week. The FTC is reportedly planning to compel executives at OpenAI and Anthropic, along with staff at the independent AI testing group METR, to testify, and intends to send similar demands to other AI developers in the coming weeks. That would widen an inquiry that, as of late September, was already understood to be examining the broader industry's safety and marketing practices.
Company Responses and What Comes Next
OpenAI said it intends to engage constructively with the FTC's inquiry. Anthropic did not immediately respond to requests for comment reported by outlets covering the story. Neither company has disputed that it received a demand.
Civil investigative demands do not themselves constitute a finding of wrongdoing, and the FTC can close such inquiries without taking further action. But the scope described by people familiar with the matter — executive testimony, safety-incident records, and marketing claims going back through both companies' histories — suggests the agency intends a sustained look at how leading AI labs have described their products' capabilities and safeguards to the public, at a moment when both companies are fielding billions of dollars in enterprise and consumer-facing deployments.