The draft policy statement argues that AI companies could violate federal consumer protection law if AI outputs intentionally veer from what users reasonably expect.
The Federal Trade Commission (FTC) is seeking public comment on a proposed policy statement explaining how federal consumer protection law could apply to AI systems. If adopted, the policy would state that AI companies violate the FTC Act when they intentionally alter AI outputs without clearly informing users. The agency argues that consumers reasonably expect AI tools to pursue user-stated objectives and provide the best answers possible.
Accuracy as a consumer protection issue
The proposed policy states that AI companies may violate the FTC Act if they intentionally steer AI outputs toward undisclosed objectives that conflict with what users requested or reasonably expected. According to the FTC, that could include modifying responses for ideological, political, or other undisclosed reasons without adequately informing consumers.
The agency distinguishes those practices from ordinary AI mistakes, sometimes called “hallucinations.” The proposal says factual errors resulting from the technical limitations of AI models are different from intentionally designing systems to prioritize other objectives over accuracy.
Disclosure would need to be clear
The FTC acknowledges that AI developers may design their systems to balance multiple goals. However, it argues that if those goals intentionally influence answers in ways users would not reasonably expect, companies should clearly disclose that to consumers.
According to the proposal, those disclosures would need to be prominent and understandable. The FTC says companies generally could not rely on language buried in terms of service or other fine print if the overall impression given to consumers is that an AI system is designed to provide the best available answer.
Potential conflict with some state AI laws
One of the proposal’s more notable points is that compliance with certain state AI laws would not necessarily protect companies from FTC enforcement if consumers are deceived.
The policy statement cites Colorado’s revised AI law as an example of a state measure that could encourage companies to steer AI responses away from what users would reasonably expect. The proposal argues that federal consumer protection law continues to apply even when companies are attempting to comply with state requirements.
Building on existing FTC AI enforcement
The FTC cites several recent enforcement actions involving AI products, including cases alleging deceptive claims about AI capabilities and performance, as examples of how the agency has already applied longstanding consumer protection principles to AI technologies.
The FTC is accepting public comments on the proposed policy statement before determining whether to issue a final version.
