FTC AI Accuracy Policy Statement

Overview

The Federal Trade Commission (FTC) has issued a Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems, signaling a critical intervention into the burgeoning field of AI. This policy statement, published on July 1, 2026, aims to clarify how Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices, applies to AI systems that may manipulate or distort outputs contrary to reasonable consumer expectations. The FTC's primary concern revolves around instances where AI companies explicitly or implicitly represent their systems as providing accurate, neutral, or objective information, yet deliberately configure them to prioritize undisclosed objectives over truthful outputs. This includes situations where AI models are trained or tuned to steer outputs away from the most accurate answer without transparently informing users. The policy distinguishes between unintentional technical limitations, such as 'hallucinations' due to a lack of knowledge, and intentional design choices to suppress accuracy, with the latter being the focus of potential enforcement actions.

The impetus for this policy statement stems from the increasing reliance of consumers on AI systems for information and decision-making, coupled with the FTC's observation that consumers often accept AI outputs without independent fact-checking. The Commission asserts that consumers have a reasonable expectation that AI systems are not secretly designed to pursue undisclosed ideological goals or other objectives that distort their outputs. This initiative aligns with the FTC's broader efforts to protect consumers from deceptive AI practices, including exaggerated performance claims, false labeling of AI-driven products, opaque data practices, and algorithmic bias. The policy statement also addresses the complex interplay between federal consumer protection laws and emerging state-level AI regulations, particularly those that might compel AI companies to alter outputs in ways that could conflict with federal deception standards. The public comment period for this proposed policy statement is open until July 31, 2026, inviting input from businesses, consumers, and other stakeholders.

Definitions

The Proposed Policy Statement, while not introducing new statutory definitions, clarifies the FTC's interpretation of existing legal concepts in the context of Artificial Intelligence (AI) systems. Central to this policy is the understanding of 'accuracy' in AI outputs, which the FTC views through the lens of consumer expectations. AI systems are broadly understood as tools capable of making predictions, recommendations, or decisions, with their utility judged by how well they match consumers' stated or reasonably expected objectives. The policy emphasizes that when companies market AI as accurate, neutral, or objective problem-solving tools, consumers form a reasonable expectation of truthful and faithful outputs.

Key terms implicitly defined by the policy's focus include 'suppression of accuracy,' which refers to the deliberate act of training or configuring an AI model to prioritize objectives other than the most accurate answer, without clear and conspicuous disclosure to the user. This is distinct from 'technical hallucinations,' which are errors arising from a model's inherent limitations or lack of knowledge, and are generally not the target of this policy. The concept of 'deceptive acts or practices,' as outlined in Section 5 of the FTC Act, is applied to AI systems where representations, omissions, or practices are likely to mislead a reasonable consumer in a material way. The policy posits that undisclosed ideological steering of AI outputs, regardless of motivation, constitutes a deceptive practice. Furthermore, 'adequate disclosure' is defined as a high bar, requiring prominence and persistence, going beyond mere disclaimers buried in terms of service or technical documentation.

Governance and Institutional Framework

The governance framework for the FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems is firmly rooted in the existing authority of the Federal Trade Commission (FTC) under Section 5 of the FTC Act. As an independent agency, the FTC is empowered to prevent unfair methods of competition and unfair or deceptive acts or practices in commerce. This policy statement serves as a crucial interpretative guidance, clarifying how these established legal principles apply to the unique challenges presented by advanced AI technologies. The FTC has historically adapted its consumer protection mandate to new technologies, from radio to the internet, and now extends this oversight to AI, focusing on consumer and competition issues. The Commission's approach to AI regulation is primarily enforcement-based, leveraging its existing statutory powers rather than establishing entirely new AI-specific regulations.

The institutional framework involves the FTC's Bureau of Consumer Protection, which is at the forefront of identifying and addressing deceptive AI practices. The agency's actions, such as the enforcement case against Rite Aid for discriminatory facial recognition technology, demonstrate its readiness to act against AI systems that cause consumer harm. The FTC also collaborates with other federal agencies, including the Department of Justice, the Consumer Financial Protection Bureau, and the Equal Employment Opportunity Commission, as evidenced by joint statements on enforcement efforts against discrimination and bias in automated systems. This collaborative approach underscores a broader governmental commitment to ensuring fairness and preventing discrimination in AI. The policy statement itself is a product of the FTC's ongoing research and horizon scanning, including studies, reports, and public engagement initiatives, to stay abreast of AI developments and their implications for consumers and competition. The public comment period for this proposed policy statement is an integral part of this framework, allowing stakeholders to provide input that will help shape the final policy.

Key Focus Areas

The FTC's Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems zeroes in on several critical areas to protect consumers from potentially deceptive AI practices. A primary focus is on the truthfulness and substantiation of claims made by AI developers and deployers. The policy asserts that companies marketing AI systems as accurate, neutral, or objective problem-solving tools must ensure their systems' actual design and outputs are consistent with these representations. The FTC is particularly concerned with instances where AI companies manipulate system behaviors to achieve undisclosed ideological objectives, rather than the outcomes consumers request or reasonably expect. This means that any deliberate configuration of an AI model to prioritize certain objectives over pure accuracy, without transparent disclosure, could be deemed a deceptive act under Section 5 of the FTC Act.

Another key area is transparency and disclosure requirements. The policy sets a high bar for such disclosures, stipulating that they must be clear, conspicuous, and adequate to genuinely inform consumers that an AI system is designed to prioritize objectives different from what users would expect. A disclaimer buried in terms of service or technical documentation would not suffice; disclosures need to be sufficiently prominent and persistent within the user experience itself. This focus aims to ensure that consumers are not misled about how an AI system behaves, especially when its outputs are intentionally steered away from maximal accuracy. Furthermore, the policy implicitly touches upon risk management by highlighting the potential for consumer harm arising from misidentified or distorted AI outputs, which can lead to severe emotional distress, reputational harm, or wrongful arrest, as seen in previous FTC enforcement actions involving facial recognition technology. The FTC's broader interest in scrutinizing deceptive claims also extends to exaggerated performance, false labeling, opaque data practices, and bias in AI decision-making systems.

Implementation Framework

The implementation framework for the FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems relies heavily on the FTC's existing enforcement mechanisms under Section 5 of the FTC Act. This policy statement, while not a standalone regulation, serves as a clear articulation of how the FTC intends to apply its authority to address deceptive practices related to AI accuracy. Companies are expected to review their AI development and deployment practices to ensure compliance with the principles outlined in the statement, particularly regarding the truthful representation of AI system capabilities and the transparent disclosure of any intentional deviations from pure accuracy. The FTC's enforcement actions typically involve investigations, consent orders, and litigation, with remedies that can include monetary penalties, technology bans, and requirements for algorithmic accountability and privacy assessments.

A critical aspect of the implementation framework is the emphasis on substantiation of claims. Companies must be able to provide evidence that claims about AI capabilities are truthful and not misleading. This includes documenting AI capability claims with supporting evidence and performance data, maintaining an audit trail of claims made in marketing materials, and linking them to technical specifications. The policy statement also implicitly encourages the development of robust internal governance structures within companies to assess and mitigate risks associated with AI accuracy and potential deception. The FTC has previously issued formal business guidance and blog posts warning companies about deceptive AI practices, and this policy statement further solidifies those warnings. The ongoing public comment period is a key part of the implementation process, allowing the FTC to gather diverse perspectives and refine its final policy, ensuring it is both effective in protecting consumers and practical for businesses.

Monitoring and Evaluation

Monitoring and evaluation of the FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems will be an ongoing process, primarily conducted through the Federal Trade Commission's (FTC) established mechanisms for market surveillance, consumer complaints, and enforcement actions. The FTC continuously monitors the marketplace for unfair or deceptive practices, and the proliferation of AI technologies necessitates a heightened focus on how these systems are marketed, used, and perceived by consumers. Consumer complaints will serve as a vital source of intelligence, alerting the Commission to potential instances where AI outputs are being suppressed or manipulated without adequate disclosure, leading to consumer harm. The FTC's ability to receive and analyze these complaints, coupled with its investigative powers, will be crucial in identifying violations of the principles outlined in the policy statement.

Furthermore, the FTC will evaluate the effectiveness of the policy through its enforcement outcomes. Successful enforcement actions against companies found to be engaging in deceptive AI practices will demonstrate the policy's impact and deterrent effect. The agency's ongoing research and horizon scanning activities, including studies and reports on AI trends, will also contribute to the evaluation process, allowing the FTC to assess the evolving landscape of AI technology and its implications for consumer protection. This adaptive approach ensures that the policy remains relevant and responsive to technological advancements. The policy statement itself, as a proposed document, is currently undergoing public comment, which is a form of initial evaluation, allowing stakeholders to provide feedback on its clarity, scope, and potential impact before it is finalized. This iterative process of engagement, enforcement, and continuous learning will enable the FTC to refine its approach to AI accuracy and consumer protection over time.

Penalties, Liability, and Appeals

The FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems clarifies the application of existing legal frameworks concerning penalties, liability, and appeals under Section 5 of the FTC Act. Companies found to be engaging in deceptive practices by suppressing AI accuracy without adequate disclosure face significant legal consequences. Violations of Section 5 can result in substantial civil penalties, which can be considerable given the per-violation, per-day nature of some penalties. Beyond monetary fines, the FTC has demonstrated a willingness to impose injunctive relief, which can include requirements for algorithmic disgorgement (destroying algorithms or data trained on unlawfully collected information), technology bans, and ongoing monitoring or auditing requirements for AI systems. For instance, the FTC banned Rite Aid from using facial recognition technology for five years after allegations of unfair deployment and lack of safeguards.

Liability for deceptive AI practices extends to companies that market AI tools with exaggerated performance claims, make misleading statements about AI training data or methods, or fail to disclose limitations or intentional alterations to AI outputs. The policy statement underscores that even efforts to comply with state AI laws that require alterations to AI outputs could lead to federal deception liability if not adequately disclosed to consumers. Companies are expected to substantiate claims about AI safety, bias, and efficacy. Appeals of FTC enforcement actions generally follow established administrative and judicial review processes. Companies can challenge FTC decisions in federal courts, where the courts review whether the FTC acted within its statutory authority and followed proper procedures. The FTC's policy statements, while influential, do not carry the same precedential weight as litigated cases or consent orders, but they strongly signal the agency's enforcement priorities and legal theories.

Relationship to Other Instruments

The FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems operates within, and significantly informs, a broader landscape of legal and policy instruments concerning artificial intelligence in the United States. It does not introduce new standalone AI-specific regulations but rather applies existing consumer protection laws, primarily Section 5 of the FTC Act, to AI systems. This approach aligns with the FTC's historical practice of adapting its core mission to emerging technologies. The policy statement complements other FTC guidance and enforcement actions, such as those addressing fake reviews, impersonation via deepfakes, AI robocalls, and the unlawful use of children's data for AI training under COPPA. It also builds upon the FTC's focus on algorithmic bias and discrimination, as articulated in joint statements with other federal agencies like the DOJ, CFPB, and EEOC, which emphasize the application of civil rights and fair competition laws to automated systems.

A notable aspect of this policy statement is its explicit consideration of the relationship between federal consumer protection mandates and state-level AI regulations. The FTC's document addresses concerns that state laws, such as the Colorado Artificial Intelligence Act, which may compel companies to steer AI outputs toward specific objectives, could inadvertently lead to deceptive practices under federal law if not properly disclosed. The policy statement suggests that such state laws might be impliedly preempted where they conflict with a federal regulatory scheme, particularly the FTC's authority to police deception. This highlights a potential area of tension and the FTC's intent to establish a clear federal stance on AI accuracy and consumer expectations. The policy also draws from broader executive orders and administrative priorities concerning AI, reflecting a national effort to balance innovation with responsible AI development and deployment.

International Alignment

While the FTC Proposed Policy Statement on the Suppression of Accuracy in Artificial Intelligence Systems is a domestic U.S. instrument, its underlying principles of transparency, accuracy, and consumer protection in AI resonate with emerging international discussions and frameworks. Many global initiatives, such as those from the OECD, G7, and various national strategies, emphasize the importance of trustworthy AI, which inherently includes aspects of accuracy, reliability, and explainability. The FTC's focus on preventing deceptive practices where AI outputs are manipulated aligns with a global consensus that AI systems should not mislead users or operate in opaque ways that undermine public trust. Although the policy statement does not explicitly detail international alignment, the FTC's broader engagement in international forums on consumer protection and competition issues provides a context for its approach to AI. The agency participates in cross-border cooperation and shares best practices with international counterparts, recognizing that AI innovation and its challenges transcend national borders.

The emphasis on clear and conspicuous disclosure regarding AI system behavior, particularly when outputs are intentionally steered away from pure accuracy, reflects a principle that is gaining traction in various jurisdictions. International guidelines often call for mechanisms to inform users about the capabilities and limitations of AI systems, and the FTC's policy contributes to this global dialogue by setting a high bar for such transparency. The challenges posed by AI, such as algorithmic bias and the potential for manipulation, are universal concerns that international bodies are actively addressing. By clarifying its stance on AI accuracy and deception, the FTC contributes to a growing body of regulatory thought that can inform and influence future international standards and recommendations for responsible AI development and deployment. The policy's consideration of preemption concerning conflicting state laws also touches upon broader international discussions about regulatory harmonization and the potential for a fragmented global AI regulatory landscape.

Implementation Timeline

The proposed policy statement was authorized by a 2-0 vote of the Federal Trade Commission on July 1, 2026, and published in the Federal Register on July 7, 2026 (91 FR 41638). The public comment period ran through July 31, 2026. Following staff review of submitted comments in late 2026, the Commission will determine whether to issue a final policy statement.

Sources and References