FTC opens probe into Anthropic, OpenAI and other AI companies

By Splitfeed EditorialLast updated Sep 30, 2026, 15:10 UTC

On September 30, 2026, the Federal Trade Commission opened what has been described as a sweeping probe into Anthropic, OpenAI, and other companies developing superintelligence models. The investigation was first reported by the New York Post and subsequently confirmed by Reuters and CNBC, among others. The same day, executives from Google, Anthropic, Meta, Nvidia, OpenAI, and xAI signed a White House accord on superintelligence, a parallel development that complicates any simple narrative about the relationship between these companies and the federal government. Separately, OpenAI, Google, and Anthropic agreed to appear at a New York City AI hearing after being threatened with subpoenas.

The FTC probe matters because it represents one of the most direct regulatory moves yet against the companies building what the industry itself is calling superintelligence. Unlike earlier AI investigations that focused on data practices or competitive behavior, this one is framed around the potential risks of the models themselves. That framing raises genuinely contested questions: what legal authority the FTC has to regulate emergent AI risks, what specific harms it is attempting to document, and whether a probe of this scope could produce enforceable outcomes or will largely serve as a pressure mechanism. The White House accord signed the same day adds another layer of complexity, since it suggests these same companies retain significant access to federal policymakers even as one agency scrutinizes them.

No reactions from the Pro-AI, Anti-AI, or Middle Ground camps have been published yet. Typically, the Pro-AI camp would argue that regulatory probes of this kind risk chilling research at a critical moment of development and that industry self-governance, reflected in the White House accord, is a more agile response to fast-moving technology. The Anti-AI camp would generally welcome the FTC action as overdue and argue that voluntary accords carry no real accountability. The Middle Ground camp would likely ask for specifics about the probe's scope and legal basis before drawing conclusions about whether it represents meaningful oversight or political signaling.

The details that would settle much of this debate include the formal scope of the FTC's investigative demands, any timeline the agency sets for responses, and what, if anything, the White House accord actually commits the signing companies to do. The New York City hearing, which OpenAI, Google, and Anthropic agreed to attend under subpoena threat, could also surface specific claims about risk that clarify what regulators believe they are dealing with.