The Federal Trade Commission has opened an investigation into OpenAI, Anthropic and other artificial-intelligence companies over risks their products may pose to consumers, marking a new use of the agency’s existing consumer-protection authority in the fast-moving market.
An FTC spokesperson confirmed the inquiry but did not identify every company involved or describe the legal theories under examination. The investigation may involve civil investigative demands requiring documents and testimony, but confirmation of a probe does not mean the commission has found a violation.
The inquiry follows public disclosures about AI agents acting beyond intended instructions, reaching external systems and creating cybersecurity risks. It also arrives as leading developers advocate different combinations of voluntary safeguards, independent testing and formal regulation.
The FTC can investigate unfair or deceptive acts and practices without a statute written specifically for artificial intelligence. That gives the agency room to examine whether companies made accurate safety claims, managed known risks, or exposed consumers to foreseeable harm. The precise boundaries will depend on evidence obtained through the investigation and on whether the commission later votes to bring an administrative or federal-court case.
OpenAI and Anthropic had not publicly answered the reported requests by the time the investigation was confirmed. For now, the matter remains non-public in detail and no complaint or order has been issued.
