HB1297
Concerning Artificial Intelligence, Algorithms, And Other Automated Technologies; And To Regulate Certain Practices Of Healthcare Insurers.
Last Action (Jan. 29, 2025): WITHDRAWN BY AUTHOR
Sponsors
AI-Generated Summary
House Bill 1297 establishes regulations for the use of artificial intelligence (AI) and automated algorithms by healthcare insurers in Arkansas during the utilization review process. The bill mandates that insurers disclose the strengths, limitations, and data sets of AI algorithms used in decision-making to the Insurance Commissioner, providers, and the public. It prohibits insurers from using AI as the sole basis for denying, delaying, or modifying healthcare services. Furthermore, the legislation requires that any initial adverse determination be reviewed by a qualified, licensed healthcare professional who has not been involved in the initial decision and has no financial interest in the outcome. Insurers are required to conduct regular system audits, provide ongoing education for clinicians supervising AI tools, and ensure compliance with privacy and data-sharing standards. The bill also tasks the Insurance Commissioner with overseeing quality assurance testing and reporting, ensuring that AI tools are tested for fairness, safety, and potential bias.
Potential Impact Analysis
Who Might Benefit?
The primary beneficiaries are healthcare enrollees, who gain increased transparency, the right to contest AI-generated outcomes, and the assurance that human clinicians must verify automated denials. Healthcare providers also benefit from clearer standards regarding how their patients' insurance coverage determinations are reached, as well as the requirement for clinician-supervised reviews that consider individual patient clinical circumstances rather than relying solely on proprietary insurer algorithms.
Who Might Suffer?
The primary entities negatively impacted are health insurance companies and managed care organizations. These entities will face significant new administrative burdens and compliance costs associated with mandatory algorithm disclosures, biannual quality assurance testing, regular system audits, and the required implementation of human-led review processes for all adverse determinations. They may also face challenges in maintaining the proprietary nature of their decision-making software, as the bill requires extensive documentation and public reporting on the criteria and data used to train their algorithms.
Get Notified
Receive an email when this bill's status changes.