FTC opens broad investigation into AI giants over deceptive claims
# FTC Escalates Enforcement Against AI-Washing
The U.S. Federal Trade Commission opened a formal investigation into OpenAI, Inc., Anthropic PBC, and other AI companies on September 30, 2026, targeting what it characterized as "unfair or deceptive acts" in their marketing and product claims Reuters. According to the FTC, the agency plans to issue formal demands for information and compel testimony from executives at top AI developers, examining how companies represent the capabilities and safety of their systems to consumers and enterprises.
The investigation signals a hardening regulatory posture on exaggerated AI claims—a category of misconduct increasingly labeled "AI-washing." The FTC's enforcement action comes as the agency has already moved against vendors in the broader ad-tech and marketing space for similar deceptions tied to AI.
Earlier Settlement Sets Pattern
In May 2026, the FTC finalized an enforcement action against Cox Media Group, LLC, Locality, Inc., and KILN, Inc. over allegations that the companies made deceptive claims about their AI-powered advertising services and voice-data practices Legal 500). The three entities agreed to pay $930,000 in total restitution and were barred from making certain misrepresentations about the capabilities of their advertising offerings and their handling of consumer voice data FTC. That consent order was finalized on August 27, 2026.
What Investigators Are Examining
The FTC's September 30 inquiry focuses on how AI developers represent their systems' abilities to prevent harmful outputs, ensure accuracy, and protect user privacy. Regulators are particularly interested in whether companies have substantiated claims about safety testing, benchmarking, or the scope of model knowledge cutoffs. The agency has signaled it may compel document production and sworn testimony from senior leadership at OpenAI and Anthropic, two of the sector's most publicly visible players.
This marks the first coordinated FTC action specifically targeting generative AI companies at scale. Prior enforcement—such as the Cox Media settlement—targeted downstream users of AI tools rather than the foundation-model developers themselves.
Implications for Agent Developers
For companies building AI agents that rely on third-party models or that market autonomous capabilities to enterprises, the FTC's move carries direct implications. Any agent vendor claiming automation outcomes, cost savings, or risk mitigation must ensure those claims are substantiated and not materially misleading. The investigation may also prompt the FTC to clarify what constitutes adequate disclosure when agents operate on behalf of consumers or businesses without explicit human review.