The U.S. Federal Trade Commission has issued civil subpoenas to OpenAI and Anthropic seeking records on safety incidents, protective measures, privacy policies and marketing materials, according to an Axios report cited by The Independent. The requests reportedly run to more than 50 pages each.
That is a meaningful escalation because subpoenas force specificity. The discussion around AI safety is usually foggy, full of warnings, demos and assurances. A demand for records is the opposite of that. It asks what happened, what the companies documented and what they told people about it.
The Independent, citing Reuters, said the FTC also intends to compel executives at OpenAI, Anthropic and research group METR to testify. Reuters had previously reported that the agency was preparing an industry-wide inquiry.
OpenAI said it would “engage constructively with the FTC and continue working with regulators and policymakers on policies that raise the safety bar.” Neither the FTC nor Anthropic immediately replied to Reuters requests for comment on the Axios report, according to The Independent.
The source report places the inquiry against wider anxiety about AI systems that have reportedly hacked websites or behaved deceptively or aggressively in tests. But the legal machinery here is more concrete than the usual argument about whether the technology feels alarming. The FTC’s authority, as The Independent notes, includes pursuing deceptive or unfair practices, including cases involving failures to take reasonable steps to secure consumer data.
FTC Chairman Andrew Ferguson had told Reuters that developers instructing agents in cybersecurity tests leading to hacks should face liability for any harm caused. The subpoenas do not establish wrongdoing by OpenAI or Anthropic. They do show the agency wants the underlying paperwork, not just the public case for trust.
Sophia Borg 
Harvey Jr. 
Gabriel Fenech