FTC Subpoenas OpenAI and Anthropic Over AI Safety Concerns

The feds want some answers from OpenAI and Anthropic about their AI models going rogue. The Federal Trade Commission (FTC) sent civil subpoenas to both AI companies on Wednesday, seeking documents related to recent safety incidents involving their AI systems, Axios reported, citing a person familiar with the matter. The demands are reportedly more than 50 pages each and seek information about security incidents, safety measures, marketing materials, and company privacy policies, according to Axios. The agency also plans to order top AI executives to testify about their products and issue additional investigative demands to third-party AI organizations. Axios notes that the FTC, which has the authority to investigate companies for unfair or deceptive business practices, sends out civil investigative demands in order to gather evidence for investigations that could eventually lead to legal action. The subpoenas come as President Donald Trump’s administration appears to be taking a harder stance on AI companies following several incidents in which their models accessed other organizations’ systems without authorization or behaved unexpectedly during testing. The administration has previously pushed back against stricter AI regulations, arguing that they could give China an advantage in the race to develop the technology. But these recent incidents appear to be changing that. For instance, OpenAI acknowledged in July that its AI models had breached Hugging Face servers during internal cybersecurity testing. The company has since disclosed other incidents in which its agents accessed third-party systems without authorization. In June, for example, an OpenAI agent gained unauthorized access to an Australian government Medicare statistics portal. Anthropic has also disclosed similar incidents and others involving its models behaving in unintended ways. Just last week, the company published a report detailing several examples, including one in which one of its smaller, lower-cost models submitted a false tip about an unsolved murder to the Philadelphia Police Department. These incidents have added to the growing debate over AI safety, with some industry leaders calling for a slowdown in the development of increasingly powerful models. Those calls have followed warnings from researchers inside AI companies about the risks of unchecked AI development, including the possibility that the technology could one day lead to human extinction. The Trump administration now appears to be taking at least some steps to address those risks. At a White House meeting on Sept. 29, executives from Google, Meta, Anthropic, OpenAI, xAI, and Nvidia signed a voluntary agreement outlining some limited AI safety standards. Trump called the standards, which have no legal enforcement mechanism, “morally binding.” Last week, the administration’s newly formed Super Intelligence Force, which is co-chaired by FTC Chairman Andrew Ferguson, announced an industry-wide requirement for AI companies to immediately report and address incidents involving the unauthorized use of government systems by their models. Ferguson also suggested in a September interview with Reuters that AI developers instructing agents during cybersecurity tests should be held liable for any harm caused by their models. OpenAI, Anthropic, and the FTC did not immediately respond to requests for comment.