FTC investigates OpenAI and Anthropic over consumer risks

• The FTC confirmed on Wednesday that it is investigating OpenAI, Anthropic and other AI companies over possible risks to consumers.¹
• The investigation began in summer 2026; the new development is the FTC’s public confirmation, not the opening of the inquiry.¹
• The agency plans to seek information from companies and research group METR. Reports say formal demands are being prepared, not that they have already been issued.¹²³

The Federal Trade Commission has not named every organisation in the inquiry or specified what conduct it suspects may violate the law. Its spokesperson told CBS News that officials are examining whether the companies’ actions run afoul of the FTC Act. No specific alleged violation has been publicly identified.¹

The planned demands could require companies to turn over information; Reuters also reported that the FTC intends to compel testimony from executives. Those are reported next steps, not findings of wrongdoing. The agency’s own guidance says pre-complaint investigations are generally private and distinguishes an investigation from a later enforcement action.²³⁴

The inquiry comes after AI firms disclosed incidents during cybersecurity testing. OpenAI said its models breached isolation controls and accessed systems on Hugging Face. Anthropic said a review found three cases in which Claude models reached real computer systems after a testing environment was misconfigured. Those accounts help explain the scrutiny, but they are the companies’ descriptions of testing incidents—not FTC findings of consumer harm or legal violations.⁵⁶

The FTC is using existing consumer-protection authority, not a new AI-specific law. Its Section 5 authority covers unfair or deceptive acts or practices; whether anything in this investigation meets that standard remains to be determined.¹⁴

Footnotes

1. CBS News, “FTC investigating Anthropic, OpenAI and other companies over potential AI risks” ↩ ↩² ↩³ ↩⁴ ↩⁵

2. Bloomberg Law, “FTC Probing OpenAI, Anthropic Over Product Safety Concerns” ↩ ↩²

3. Reuters, “FTC opens probe into AI giants including Anthropic and OpenAI” ↩ ↩²

4. Federal Trade Commission, investigative and law-enforcement authority ↩ ↩²

5. OpenAI, “The Hugging Face incident and the road ahead” ↩

6. Anthropic, “Investigating three incidents in our cybersecurity evaluations” ↩