AI Frontier Institute

Relates to the use of artificial intelligence by insurers and clinical peer reviewers for utilization review; establishes additional notice requirements for adverse determinations.

View primary source →

New York Senate Bill 10241, introduced May 8, 2026 by Sen. Hinchey, amends the Insurance Law and Public Health Law to regulate insurer use of artificial intelligence (AI) in utilization review. The bill adds a new Section 338 to the Insurance Law requiring insurers to provide written notice to insureds, enrollees, and health care providers when AI-based algorithms are first adopted and once per policy period thereafter, plus a conspicuous website notice. Under Section 338(b)(2), insurers must disclose to health care providers—and to insureds upon request—the criteria governing the algorithm, its training data sets, the algorithm itself, and its outputs. Section 338(b)(4) prohibits adverse determinations based solely on group data sets, and Section 338(b)(5) prohibits algorithms from directly or indirectly causing harm to insureds. Section 338(b)(6) requires algorithms to be open to superintendent audit. Section 338(c) requires insurers to submit algorithms and training data to the superintendent with certifications that they minimize bias across protected characteristics, adhere to evidence-based clinical guidelines, and do not independently create or change clinical standards. Section 338(d)(1) mandates that adverse determinations be made only by clinical peer reviewers—not AI alone—who must individually review medical history and clinical records, and Section 338(d)(3) explicitly prohibits AI from being the sole basis for denying, delaying, or modifying health care services. Section 338(e) requires insurers to establish ongoing quality assurance testing of AI algorithms, submit results to the superintendent at least semiannually, and publish results publicly within 30 days. Penalties under Section 338(f) include license suspension or revocation, fines up to $5,000 per violation ($10,000 for willful violations), capped at $500,000 per insurer and $100,000 per clinical peer reviewer annually. Sections 8 and 9 amend Section 4903 of the Insurance Law and Public Health Law respectively to require adverse determination notices to include a signed clinical peer reviewer statement disclosing credentials, time spent, whether actual or summary records were reviewed, and a plain-language statement of any AI use. Sections 2–7 make conforming clarifications to 'emergency condition' definitions across multiple insurance and public health law provisions. The act takes effect 60 days after enactment.

← Back to the tracker