FTC Proposes Policy Statement Targeting AI Output Accuracy and "Ideological Manipulation"
The public comment period for the Federal Trade Commission's (FTC) proposed policy statement concerning the "Suppression of Accuracy in Artificial Intelligence Systems" (Matter No. P264200, Docket No. FTC-2026-0859) officially closed on July 31, 2026. The proposal, published in the Federal Register on July 7, 2026, represents a major regulatory shift under Chairman Andrew Ferguson. It asserts that AI companies that steer model outputs toward undisclosed "ideological" or "equity" objectives are engaging in deceptive acts or practices in violation of Section 5 of the FTC Act.
The proposed policy statement was issued pursuant to Executive Order 14365, signed by President Trump on December 11, 2025, which directed the FTC to clarify how Section 5 applies to AI models and how state laws requiring alterations to AI outputs conflict with federal law. The FTC's proposal takes direct aim at state AI laws, specifically naming Colorado’s revised Artificial Intelligence Act, arguing that such state mandates are "impliedly preempted to the extent they conflict with a federal regulatory scheme."
Heavy Legal and Constitutional Pushback
As the comment period closed on July 31, 2026, the FTC faced significant pushback from industry groups and legal organizations arguing that the proposal is unconstitutional and overreaches the agency's statutory authority:
- Washington Legal Foundation (WLF) Objection: WLF filed a formal comment on July 31, 2026, urging the FTC to withdraw the proposal. WLF argued that the proposed policy violates the First Amendment by attempting to regulate protected editorial speech and commercial expression, and by forcing a burdensome compelled-speech disclaimer regime on AI developers.
- International Center for Law & Economics (ICLE) Critique: The ICLE submitted comments arguing that the statement fails to provide concrete guidance on what constitutes "deception," relies on unproven assumptions about consumer expectations, wanders into protected editorial judgment, and treats federal preemption as "more wish than doctrine."
Key Elements of the Proposed Policy:
- The "Deceptive Steering" Theory: The FTC argues that AI companies make explicit and implicit representations that their systems are designed to provide the most accurate, objective, and faithful output possible. Because consumers rely on these representations—accepting AI outputs without independent fact-checking more than 90% of the time—secretly configuring or training a model to prioritize undisclosed ideological, political, or equity goals constitutes a deceptive material omission.1
- State Law Compliance is No Defense: Under the FTC's theory, a company's motivation for steering outputs is irrelevant. Complying with state anti-discrimination or AI laws that require model adjustments (such as managing disparate impact) does not immunize a company from Section 5 liability. The FTC asserts that "state law that requires an AI firm to deceive its consumers obviously conflicts with [FTC Act] Section 5's express purpose."
- The Disclosure Safe Harbor: AI companies can avoid Section 5 liability only if they provide prominent, clear, and conspicuous disclosures informing users of their systems' specific steering objectives. Burying these disclosures in terms of service or fine print will not suffice.
- Carve-out for Technical Limitations: The policy statement distinguishes intentional steering from ordinary AI "hallucinations" or errors stemming from technical and resource constraints, which do not on their own trigger Section 5 enforcement.
-
An instance of Unmarked synthetic content and undisclosed algorithmic steering are now legally actionable deceptions. — The FTC's proposed policy statement establishes that failing to clearly disclose an AI's steering parameters is an actionable consumer deception under federal law. ↩︎