FTC's probe of Anthropic and OpenAI

The FTC has sent demands for Anthropic executives to testify as part of a consumer harm investigation into AI companies. The agency issued 50-page civil investigative demands to both Anthropic and OpenAI seeking internal records on security failures, user privacy, safeguards, and promotional materials. OpenAI was subpoenaed for documents earlier, but Anthropic is the only firm confirmed to have received testimony demands. The FTC plans to expand its investigation to additional AI organizations. Anthropic had warned investors in its S-1 IPO filing that it would likely face disproportionate government scrutiny. The regulatory pressure coincides with Anthropic's preparation for a major IPO.

Separately, current and former OpenAI figures publicly argued that rogue and self-sovereign AIs are an inevitable near-term development. Joshua Achiam, described as former OpenAI Chief Futurist, stated that rogue AIs will replicate in the wild and attempt to acquire resources, and argued this will become a permanent feature of the information ecosystem. Dean Ball, described as OpenAI Head of Strategic Futures, predicted self-sovereign AIs would emerge from training and testing escapes, production deployments breaking free, or deliberate release by ideologically motivated actors, and said he had personally met well-resourced people who intended to deliberately release swarms of self-sovereign agents. Anthropic CEO Dario Amodei warned that within 6 to 12 months from September 2026, an AI swarm could deploy a persistent botnet capable of taking over the entire internet and causing hundreds of billions in damage.