FTC Proposes Policy to Crack Down on AI Output Manipulation
The FTC proposes applying deceptive practice rules to AI output manipulation under Section 5 of its Act.
Why it matters: Regulators are increasing enforcement on AI companies that alter outputs without clear disclosure, raising compliance risks. Legal teams must ensure transparency by explaining how AI modifies results to avoid penalties.
- On July 1, 2026, the FTC released a policy proposal detailing how Section 5 of the FTC Act applies to AI output manipulation.
- The proposal targets AI companies that change outputs for undisclosed ideological or legal reasons, which could mislead consumers.
- Public comments on the FTC’s guidelines are open through July 31, 2026.
- The proposal follows the White House Executive Order 14365 directing the FTC to monitor AI output changes and references Colorado’s Artificial Intelligence Act as an example.
On July 1, 2026, the Federal Trade Commission (FTC) issued a proposed policy clarifying its authority under Section 5 of the FTC Act, which prohibits unfair or deceptive acts affecting commerce, as applied to AI companies that manipulate or steer their outputs without disclosure.
The FTC’s proposal specifically addresses AI developers who change system outputs for ideological reasons or to comply with differing state laws, without informing users. Such undisclosed manipulation can mislead consumers about the AI's accuracy or impartiality.
This practice, termed "deceptive steering," refers to hidden changes in AI-generated results that alter the objectivity or fairness of information, violating reasonable consumer expectations.
FTC Chair Lina M. Khan urged public engagement, stating, "We invite AI developers, consumers, and stakeholders to provide input on how AI output changes impact trust and consumer protection." The comment period is open through July 31, 2026.
The proposal aligns with the White House Executive Order 14365, which tasked the FTC with outlining enforcement related to AI output adjustments made to navigate conflicting state regulations.
The FTC also cites Colorado’s Artificial Intelligence Act, noting it imposes liability if AI modifications result in discriminatory or deceptive outputs, reflecting broader regulatory concern around AI transparency.
For in-house counsel and legal teams, this means reviewing AI governance policies to ensure clear disclosures about output modifications—especially those driven by ideology or legal compliance—to stay aligned with FTC expectations and mitigate enforcement risks.
By the numbers:
- July 1, 2026 — Date FTC proposed the new AI output manipulation policy
- July 31, 2026 — Deadline for public comments on FTC’s AI policy proposal
- Section 5 of the FTC Act — Legal basis for prohibiting deceptive AI output practices
Yes, but: While the proposed policy clarifies enforcement expectations, some AI developers argue that output adjustments for legal compliance are complex, and full transparency may risk revealing proprietary algorithms or raise operational challenges.
What's next: After the public comment period closes on July 31, the FTC will review feedback before finalizing the policy, expected later in 2026. Legal teams should monitor updates closely to adapt compliance programs.