The Center for Democracy & Technology submitted comments urging the Federal Trade Commission (FTC) to withdraw its proposed policy statement regarding the “suppression of accuracy” in artificial intelligence systems. The FTC’s proposed framework wrongly treats standard technical steps that are needed to train AI models, and ensure they are accurate, reliable, fair, and safe, as potential “deceptive steering” away from an assumed “neutral” baseline. In reality, no purely neutral AI system exists; every design choice made by developers of AI models shapes the outputs of AI systems, sometimes in complex and unpredictable ways. By mislabeling routine design choices, accuracy-enhancing refinements, and safeguards as undisclosed “ideological bias,” the FTC’s proposed policy threatens to penalize companies for following industry best practices.
Beyond its technical flaws, the policy statement represents significant constitutional and statutory overreach. An AI developer’s design decisions about training data, model tuning, and generation policies constitute protected editorial judgment under the First Amendment, much like traditional curation by media platforms. By attempting to police the outputs of AI models and forcing developers to characterize how their editorial choices diverge from the government’s baseline, the FTC crosses into impermissible content- and viewpoint-based speech regulation. Furthermore, the policy stretches Section 5 of the FTC Act beyond its statutory bounds, suggesting broad implied claims of “neutrality” without identifying concrete misrepresentations, proving reasonable consumer reliance, or establishing valid federal preemption over state laws.
The proposed policy statement would not make AI systems more accurate or objective. Instead, it would create legal uncertainty that chills developers from implementing essential safeguards against fraud, self-harm, and discrimination, and enable the FTC to inappropriately pressure AI developers to tune their models to reflect the government’s political and social ideology. While the FTC plays a crucial role in policing genuinely deceptive AI practices, such as false claims about an AI product’s functionality, it should not misuse consumer protection doctrine to police private expression.
Coalition Urges Senate Not to Let Companies Waive Financial Regulations for AI
CDT joined AI Now Institute, American Civil Liberties Union, and several organizations dedicated to tech policy, consumer protection, and civil rights in a letter to Senate leadership and the Senate Banking, Housing, and Urban Affairs Committee opposing the “AI Innovation Labs” language in Sec. 10509 of the CLARITY Act.
As concern about risks and harms related to AI systems continue to grow, a growing chorus of policymakers, industry leaders, and advocates have called for independent AI assessments. This explainer provides an overview of recent proposals for third-party assessment in the United States, including state and federal legislation, executive actions, and industry proposals.
Having third parties assess AI systems might seem like common sense, but crafting effective policies toward this goal can be devilishly tricky. A poorly-constructed ecosystem for third-party assessment could easily fail to consider the most consequential mechanisms of risk, neglect the AI harms that most impact people, or do more to protect AI companies than people.
Not All Guardrails Are Created Equal: Comparing Content Safety and Copyright Filtering
As courts and policymakers work through questions about chatbot liability, they should be wary of analogies that flatten meaningful technical differences. Copyright filtering and safety intervention share real challenges around ambiguity and evasion, but they diverge in what each control must assess, how each manifests over the course of a conversation, and how much can be verified from the outside.
