The agency said chatbot responses shaped by “ideological objectives” and some efforts to avoid discriminatory outputs could trigger scrutiny under Section 5 of the FTC Act.
The Federal Trade Commission said AI companies may violate federal law if their chatbots generate responses reflecting “ideological objectives,” outlining a proposed policy for how it plans to apply its authority to the sector. The agency said training chatbots to avoid responses that discriminate against specific groups of people could run afoul of Section 5 of the Federal Trade Act, which bars unfair or deceptive business practices, and added that complying with a Colorado law aimed at preventing AI-driven discrimination in employment and other consequential decisions could also violate the FTC Act. The move adds legal uncertainty for developers trying to limit algorithmic bias while managing U.S. consumer protection risk, and comes amid complaints from U.S. President Donald Trump and other conservatives that chatbots are politically biased against them. The FTC will accept public comment on the proposed policy until July 31.