The Federal Trade Commission on Wednesday said two common AI safety practices could violate federal law: training chatbots to avoid discriminatory responses may run afoul of Section 5 of the Federal Trade Act, and producing responses that reflect "ideological objectives" could also be unlawful. The proposed policy statement, according to the agency, means that complying with Colorado's AI anti-discrimination law could itself be deemed an unfair or deceptive practice.
The FTC will accept public comments on the proposal through July 31. Chairman Andrew Ferguson has previously used the agency's Section 5 authority in cases addressing conservative grievances, including an action against a transgender health nonprofit.
Colorado anti-bias law caught in the crossfire
Colorado's AI law aims to prevent algorithmic discrimination in employment, housing and financial services. The FTC said that following that state mandate could trigger a federal enforcement action. This immediate conflict forces companies operating across multiple states to choose between conflicting legal obligations.
Political pressure behind the policy
President Donald Trump and other conservatives have long accused AI chatbots of political bias. The FTC's latest move extends a pattern of conservative officials using federal authority to challenge what they view as anti-conservative slant in technology platforms.
Developers silent as public comment opens
Anthropic and OpenAI did not immediately respond to requests for comment. Alphabet's Google declined to comment. Stakeholders have until July 31 to submit feedback before the policy is finalized.
Why this matters for legal professionals
For lawyers advising on AI governance, the FTC's proposal introduces a direct clash between state anti-discrimination requirements and federal unfair-practice standards. In-house counsel and compliance teams must now assess whether bias-mitigation efforts could expose their companies to federal liability. AI for Legal training provides practical guidance on navigating such overlapping regulatory frameworks. Government attorneys monitoring the agency's direction may also benefit from AI for Government resources that cover federal AI policy developments.
Your membership also unlocks:
