EFF Challenges FTC AI Accuracy Policy as Unconstitutional

The Electronic Frontier Foundation has joined two other digital rights organizations in opposing a Federal Trade Commission proposal that would give the agency authority to regulate accuracy in artificial intelligence systems.

In comments filed with the FTC, EFF, Public Knowledge, and Fight for the Future argue the proposed policy statement violates constitutional protections, exceeds the Commission's legal mandate, and would enable government pressure on private companies to censor speech.

Why it matters

The controversy illustrates a fundamental tension in AI governance: how to address concerns about bias and misinformation without empowering government agencies to become arbiters of truth. For companies developing or deploying AI systems, the outcome could determine whether they face conflicting mandates from federal regulators and state legislatures—and whether their editorial decisions about model outputs become subject to federal enforcement actions based on political considerations.

Three constitutional and legal challenges

The civil liberties coalition identified three critical flaws in the FTC's July proposal, which builds on the Trump administration's executive order titled "Preventing Woke AI in the Federal Government."

First, the groups argue the policy violates First Amendment protections by positioning the FTC as a judge of which AI outputs meet an undefined accuracy standard. This constitutes prior restraint on speech, they contend, with the government installing itself as an arbiter of truth—a role courts have consistently rejected.

Second, the proposal claims federal authority to preempt state AI regulations, which the organizations say exceeds the FTC's legal powers. The policy specifically targets Colorado's automated decision-making law, which applies when AI systems are used for consequential decisions around employment, housing, healthcare, and insurance. The groups note that Congress has not delegated the FTC authority to override state laws in this domain.

Third, the vague language in the proposal creates conditions for improper "jawboning"—government pressure on private entities to censor speech. The organizations warn this would lead companies to suppress content based on the administration's interpretation of bias, even as the FTC itself acknowledges difficulty in defining where bias begins.

What the FTC should focus on instead

The joint comments conclude by urging the Commission to concentrate on its established mission: promoting structural market competition and protecting consumers from genuinely unfair and deceptive practices in the AI industry and broader technology marketplace.

The organizations emphasize they have consistently supported FTC action against actual deception in technology markets while opposing government efforts to dictate permissible speech by private actors.

The full text of the comments is available on EFF's website, as first reported by the Electronic Frontier Foundation.