FTC targets AI output manipulation as deceptive conduct
FTC Takes Regulatory Aim at Covert AI Output Steering
The Federal Trade Commission issued a proposed policy statement in July 2026 addressing what it characterizes as deceptive manipulation of AI system outputs by AI companies the FTC. The Commission voted 2-0 to authorize the statement, which was published in the *Federal Register* with a public comment period open until July 31, 2026 the FTC.
The regulatory action frames the issue as a consumer protection matter under Section 5 of the FTC Act. According to the Commission's press release, "The Federal Trade Commission is seeking public comment on a proposed policy statement addressing concerns that AI companies may be manipulating the behavior of their AI systems contrary to reasonable consumer expectations for objectivity and accuracy" the FTC.
Section 5 Deception Framework
The FTC's legal theory targets what it calls "stealthy steering"—instances where AI companies distort system outputs to achieve undisclosed ideological objectives while marketing those systems as factually reliable National Law Review. The agency asserts this conduct violates the prohibition on unfair or deceptive business practices embedded in Section 5 of the FTC Act.
The proposed statement does not impose specific penalties, fines, or enforcement deadlines. Rather, it serves as guidance clarifying what the FTC considers deceptive under existing law. No violations have yet been adjudicated, and no companies have been named or sanctioned for this specific conduct National Law Review.
Colorado AI Law and Preemption
The FTC's proposal also addresses Colorado's Artificial Intelligence Act, which the agency claims may coerce AI companies to alter model outputs. In its policy statement, the Commission argues that such a state law is "impliedly preempted to the extent it conflicts with a federal regulatory scheme" the FTC. This signals potential federal-state tension over AI governance.
The three-week comment period reflects standard FTC rulemaking procedure and allows stakeholders—including AI vendors, consumer advocates, and state regulators—to submit written views before the agency finalizes its position. The July 2026 release date positions the statement as an early intervention in an emerging enforcement area, preceding any specific litigation or settlement action.
