Federal Trade Commission (FTC) Chair Andrew Ferguson is a dramatic improvement over his Biden-era predecessor, Lina Khan. However, there is one significant area where he may be even worse than Khan: the use of the FTC’s power to enforce federal antitrust and consumer protection laws to violate the First Amendment rights of American technology companies. The latest target of Chair Ferguson’s anti-free-speech crusade is artificial intelligence (AI).
On July 1st, the FTC issued a proposed policy statement entitled, “Supervision of Accuracy in Artificial Intelligence Systems.” The statement was in response to President Trump’s December 2025 Executive Order instructing the FTC to issue a policy directive explaining how Section 5 of the the FTC Act applies to AI. As the title suggests, the FTC’s policy statement suggests that AI developers are violating Section 5’s prohibition on “deceptive acts or practices” when they market their AI products as objective and unbiased, but then steer their users towards certain results that fit the ideological bias of the company’s management and/or developers.
The FTC’s attempt to regulate AI to prevent ideological bias flies in the face of Supreme Court precedents. One is Brown, et al. v. Entertainment Merchants Assn. et al., 564 U.S. 786 (2011). In this case, the Supreme Court invalidated a California law prohibiting the sale of video games to minors. The Court agreed that California had a legitimate interest in protecting young children from being exposed to violence, but that interest did not justify violating the rights of everyone under 18.
When it comes to the FTC’s efforts to police AI for ideological bias, the issue is not just that the agency’s actions would violate the First Amendment, but their motives for doing so. The government has no more authority to regulate AI for ideological bias than it does to regulate newspapers or web sites for the same reason. A more recent Supreme Court case supporting the position that the FTC’s plan to monitor AI for ideological bias is unconstitutional is Moody v. NetChoice, LLC, 603 U.S. 2024. In this case, the High Court struck down laws in Texas and Florida regulating social media’s content moderation policies. The Court ruled that social media companies had a First Amendment right to exercise editorial control over their platforms.
Even if the courts decide that the First Amendment does not prohibit the FTC from using its power to prevent deceptive acts, this is still unwise policy. First, it is not easy to define ideological bias or determine if what appears to be bias is the result of human error. As the Cato Institute’s David Inserra, Jennifer Huddleston, and Juan Londoño pointed out, “the FTC is trying to judge AI models’ accuracy and performance—two largely subjective variables—in the same way it evaluates dietary supplements’ medical-benefit claims or users being charged fees without proper notice or consent.”
In order to protect the public from being deceived by AI systems that claim to be objective, but are in fact biased, the FTC would have to monitor the training of AI systems by developers. This training consists of teaching the systems how to recognize patterns, make decisions, and solve problems. The agency would also have to monitor the inputs that shape the algorithms used in AI searches. The agency would also have to define ideological bias. This is not as easy as it appears.
For example, a search on U.S. foreign policy that seemed to rely on sources critical of the U.S. may seem biased toward progressive Democrats, but it could also be bias toward a libertarian or Tucker Carlson-style America First version of right-wing populism. Similarly, a post favoring the point of view that the American people will never accept woke socialism could be a result of right-wing Republican bias or a bias toward center-left Democrats. Forcing developers to spend their time guarding against ideological bias could harm consumers by diverting resources away from improving AI.
Chair Ferguson and his supporters have also failed to consider what the fate of this new AI policy will be when the next Democratic president appoints Lina Khan 2.0 to head the agency. At that point, the FTC’s power to act as the AI bias police will be turned against right-wing sources under the guise of fighting misinformation and hate. The FTC’s policy guidance is a solution in search of a problem. If there is ideological bias in AI applications, entrepreneurs will develop improved AI systems that minimize bias (since those are systems developed by humans, they are not likely to be 100% free of bias). The solution to any issues users have with AI lies in the real market.
