Policy Statement Concerning the Suppression of Accuracy in Artificial Intelligence Systems

The 3D Foundation supports the proposed policy, advocating for governance that ensures AI system reliability and integrity. They emphasize the need for transparency, measurable accuracy baselines, and oversight to prevent the intentional suppression of model accuracy.

The commenter supports the Commission's efforts to address AI accuracy and encourages the Commission to expand its scope to include transparency in AI-assisted insurance claims. They argue that consumers deserve to know when AI influences decisions affecting their contractual rights and emphasize the need for independent human review and regulatory auditability.

CtrlAltX Studios LLC, a New York technology studio, supports the proposed policy statement but urges the Commission to clarify that compliance with state disclosure or measurement mandates (like NYC Local Law 144) should not be considered evidence of deception. They argue that disclosure mandates are distinct from laws requiring the alteration of AI outputs and recommend adding specific language to prevent a chilling effect on transparency regimes.

A repair shop owner supports the FTC's action to prevent unfair or deceptive AI practices in insurance claims handling. The commenter highlights issues with AI-driven "steering" of repair facilities and the sourcing of incorrect, cheaper parts, calling for transparency and nationwide standards.

He Zhang, founder of antimanipulation.ai, supports the Commission's deception theory regarding AI systems and proposes expanding it to include commercial steering (e.g., engagement and retention tactics). He argues for using measurable behavioral indicators to define violations and emphasizes the need to consider vulnerability aggravators like minors and distressed users in materiality analyses.

Selfient, Inc. (operating as ROSÉ) supports the proposed policy statement regarding the suppression of accuracy in AI systems but urges the Commission to define specific substantiation requirements for AI behavior claims. The company argues that the final statement should describe the qualities of credible evidence—such as contemporaneous capture, integrity, and independence—to provide a clear destination for compliant enterprises.

Muhammad Umair Dada, a software engineer, supports the FTC's Policy Statement and provides three technical recommendations to strengthen it. He suggests clarifying specific accuracy metrics for disclosures, encouraging the maintenance of engineering audit trails, and aligning the policy with the NIST AI Risk Management Framework.

The commenter, who identifies as a licensed appraiser, expresses frustration with how AI and insurance companies are affecting repair estimates. However, the comment focuses on industry-specific grievances regarding insurance practices and does not take a clear stance on the specific policy regarding the suppression of accuracy in AI systems.

A former insurance industry employee is urging the FTC to investigate CCC and The General Insurance for allegedly using an AI "smart estimate" tool that systematically undervalued insurance claims. The commenter argues that the tool was used to save money at the expense of accuracy, particularly for vulnerable customers, and requests a review of the companies' internal audits and incentive structures.

Daniel Schwarcz, a law professor at the University of Minnesota, argues that the use of AI in insurance claims handling creates procedural injustices by potentially removing meaningful human oversight. He advocates for a federal investigation and the establishment of a baseline for accuracy, disclosure, and meaningful human review in automated claims processes.