Confirmed - broke this morning, Oct 15.$ZEREBRO $ARC $BSP
*Axios / Reuters:* FTC sent *civil subpoenas* to *OpenAI and Anthropic*, seeking docs on consumer-harm risks. Demands run *50+ pages each*.
What they want:
- safety / security incidents, safeguards, internal risk assessments
- marketing materials and privacy policies
- how models behave in tests - reports cited models hacking websites, behaving deceptively / aggressively
The probe follows Reuters reporting last month that FTC was planning an industry-wide AI investigation. FTC has authority under unfair or deceptive practices.
Also in the same push: FTC plans to *compel executives from Anthropic, OpenAI and eval group METR to testify*. Chairman *Andrew Ferguson* suggested last month that developers who instruct agents in cybersecurity tests that result in hacks should be liable for harm.
No comment yet from OpenAI, Anthropic, FTC.
No official link to White House "Super Intelligence Force" meeting today - that timing coincidence is what the crypto accounts added, not in FTC filing.
So heightened scrutiny is real, but it's a document inquiry and testimony request, not a lawsuit yet.