The Federal Trade Commission recently issued a proposed policy statement that could change the way financial services companies use AI-powered tools. The proposed statement, entitled “Proposed Policy Statement on Suppressing Accuracy in Artificial Intelligence Systems,” alerts AI developers that changing AI outputs to deviate from accuracy, even if done to comply with a state law, may constitute consumer deception under Section 5 of the FTC Act.
The Core Theory: Accuracy Suppression as Deception
In December 2025, the government issued an executive order requiring the FTC to issue a policy statement applying the FTC Act’s prohibition on unfair and deceptive acts or practices to AI models. Based on this request, the Commission has alleged that AI companies that “direct the outputs of their AI systems toward unexpected goals and deviate from the goals set or reasonably expected by users are likely to deceive consumers in violation of Section 5 of the FTC Act.” While the use of AI tools in financial services may be new, the FTC applies its tried-and-tested three-part deception test from its 1983 policy statement to modern technology: First, there must be a representation, omission, or practice that is likely to mislead the consumer; Secondly, the consumer must act reasonably under the circumstances. and third, the representation, omission or practice must be so material that it is “likely to influence the consumer’s behavior or decision with respect to a product or service.”
AI companies market that their systems aim to produce results that achieve users’ goals as faithfully and accurately as possible. The FTC states that these representations are material “without appropriate disclaimers or qualifications” because consumers rely on them to evaluate whether a particular system is best suited to achieve their objectives. The Commission raised concerns that consumers could be tricked into paying for a service that does not behave as advertised or that they could be tricked into relying on technology that delivers inferior results if “the AI developer’s hidden agenda undermines consumers’ goals.”
The State Law Dimension: Colorado and Implicit Preemption
The proposed statement specifically discusses the interplay between the development of AI systems and compliance with government laws. Colorado’s original artificial intelligence law, for example, imposed a broad duty on AI companies to avoid outcomes that could lead to disparate impacts. The revised version of this law, recently passed in May 2026 and previously discussed here, specifically provides that AI companies can be held liable for discriminatory results caused by their customers’ use of their products. The FTC recognizes that to avoid liability under this law, AI companies may be motivated to suppress accuracy to provide fairer results. However, based on the proposed declaration, companies using AI tools could be liable for deceptive practices under Section 5 if they fail to disclose how embedding state law compliance into their models may impact accuracy.
The FTC takes the position that compliance with a state law is not a defense to a violation of Section 5. The statement says, “A business’s motives for deceiving consumers are irrelevant to the application of Section 5,” and state law is “implicitly excluded to the extent it conflicts with a federal regulatory regime.” This implicit preemption argument could have far-reaching implications for the growing patchwork of state laws on AI.
Why this matters for consumer finance
Consumer finance professionals should pay close attention for several reasons.
The adoption of AI in financial services is accelerating.
Consumers are increasingly turning to AI for financial advice, and AI tools are being integrated into lending, underwriting, and customer-facing applications across the industry. The language of consumer expectations in the proposed statement, particularly the finding that consumers accept AI outputs without further fact-checking over 90% of the time, underscores the extent to which consumers rely on AI outputs and the importance of disclosures regarding the basis of recommendations and other advice.
Fair lending and compliance with anti-discrimination regulations may be affected.
Many financial institutions have developed AI governance frameworks to identify and mitigate disparate impacts in lending and lending decisions. The proposed statement’s skepticism of laws requiring AI companies to avoid disparate outcomes raises questions about how the FTC will evaluate bias mitigation measures in AI-driven financial products. Financial institutions and their AI providers must carefully consider whether their bias testing and output adjustment practices could be classified as “suppression of accuracy” within this framework.
Disclosure requirements are being tightened.
The FTC recognizes that AI companies can influence consumer expectations through clear and conspicuous disclosure of the goals of their systems. However, disclaimers must not be buried under the lens of service and must clearly dispel the notion that the system is designed to provide the most accurate answer. For consumer-facing AI products in financial services, this means that disclosures about model limitations, performance adjustments or design decisions must be at the forefront. Furthermore, the more the disclosure contradicts the legitimate expectations of consumers, the more sustained and clear it must be.
Hallucinations are treated differently.
The FTC distinguishes between accuracy suppression due to design decisions and false outputs due to technological limitations (e.g., hallucinations). The FTC does not believe that hallucinations “in and of themselves” raise Section 5 issues, although misrepresentation of the likelihood of hallucinations could still constitute deception. This view differs from previous CFPB guidance on the use of chatbots in consumer finance. In 2023, the CFPB warned of the serious consequences that can arise if a person relies on a chatbot that relies on unreliable technology or inaccurate data. “In cases where financial institutions rely on chatbots to provide people with certain information that is required by law to be accurate, inaccuracy may violate those legal obligations.” While these concerns are not entirely resolved, we believe this departure is consistent with the current administration’s cautious optimism about advancing AI integration in financial services.
I’m looking forward to
The public comment period ends July 31, 2026. Given the FTC’s direction, financial institutions and their AI providers should begin evaluating their AI governance frameworks to ensure that any spending adjustments are accompanied by appropriate, clear disclosures and are defensible under the FTC’s accuracy-focused lens.
https://www.consumerfinancemonitor.com/2026/07/14/ftc-takes-aim-at-ai-accuracy/
