The US Federal Trade Commission is widening its investigation into Anthropic, OpenAI, and other AI companies following recent hacking incidents. The regulator will demand information and testimony from top executives. The probe also targets MET R, an AI safety research group accused of ties to Anthropic's investors.
This marks the first official regulatory action focused on rogue AI agents amid industry risks. The investigation expansion follows a White House meeting where President Trump gathered AI executives. Anthropic, OpenAI, and four other labs subsequently signed a self-regulation accord to monitor cybersecurity and biosecurity threats.
Despite these measures, the White House maintains existing laws are sufficient to limit harms from frontier models. The FTC could potentially use violations of the accord as evidence in deceptive practices cases. Meanwhile, OpenAI CEO Sam Altman missed a congressional hearing on rogue AI agents invited by Senator Josh Hawley.
Florida's attorney-general also sued OpenAI and moved to halt advanced model development until third-party safety guardrails are implemented.
Source: Financial Times Companies · Summarized by HeadlinesBriefing