Health Insurance - Artificial Intelligence - Grievance Process and Reporting (AI Health Insurance Accountability Act of 2026)
HB0795, the AI Health Insurance Accountability Act of 2026, would regulate how Maryland health carriers use artificial intelligence, algorithms, and other software tools in utilization review and grievance handling. The bill requires carriers’ internal grievance processes to provide human review when an adverse decision was made using AI or similar tools, and it ties that review to existing utilization review standards in the Insurance Article. It also directs carriers to report detailed information about grievances and adverse decisions involving AI, including the type of claim, policy type, and certain demographic categories, and requires additional reporting if AI-driven adverse decisions generate a high rate of grievances.
The bill further establishes substantive standards for AI use in health coverage decisions. Carriers, pharmacy benefits managers, and private review agents using AI for utilization review would have to ensure decisions are based on individual clinical information, not solely group datasets, and that AI does not replace the role of a health care provider. The bill also prohibits AI tools from denying, delaying, or modifying health care services, requires quarterly review of performance and outcomes, and authorizes the Insurance Commissioner to inspect AI tools and use reported information for examinations. The measure would take effect October 1, 2026.
HB0795 would amend Maryland Insurance Article §§ 15–10A–02 and 15–10A–06 and add new § 15–10B–05.1, creating explicit statutory limits and oversight requirements for AI-assisted utilization review in health insurance. It would expand carrier grievance procedures to require human review of AI-related adverse decisions, increase reporting obligations to the Insurance Commissioner, and impose compliance standards on carriers, pharmacy benefits managers, and private review agents that use AI or similar software tools. The bill would also give the Commissioner additional data for monitoring trends, potential discrimination, and model performance, and could support enforcement or examinations under existing insurance oversight authority.
The bill appears to have been introduced as a consumer-protection and accountability measure focused on transparency, human oversight, and limits on automated health coverage decisions. Because there are no committee transcripts or recorded votes in the provided materials, there is no documented floor or committee debate to gauge broader legislative sentiment. The only recorded action is that the bill was withdrawn by the sponsor in the House, suggesting it did not advance to a vote and may not have received a formal committee recommendation.
The main points of contention likely center on the bill’s restrictions on AI in utilization review, especially the prohibition on AI tools denying, delaying, or modifying services and the requirement for human review of AI-based adverse decisions. Insurers, pharmacy benefits managers, and review agents may view the reporting, auditing, and model-review requirements as burdensome or as limiting the use of automated decision-support tools, while supporters would likely argue that these safeguards are necessary to prevent unfair denials, discrimination, and overreliance on algorithms. The bill also raises privacy and data-governance issues through its reporting of demographic information and its requirements regarding patient data use.