U.S. Insurers Face Divergent Regulatory Approaches on AI Governance

Regulatory bodies in the U.S. insurance sector are divided on advancing a model law to regulate artificial intelligence (AI) use by insurers.

The National Association of Insurance Commissioners (NAIC) has taken preliminary steps, adopting AI principles in 2020 and a model bulletin in late 2023, with adoption by 24 states so far. However, some regulators urge more proactive regulation to keep pace with rapid AI adoption in insurance. Colorado has already implemented binding rules for AI governance in insurance, requiring life insurers to establish risk management frameworks for external consumer data use.

Conversely, other regulators advocate patience, emphasizing the need to thoroughly understand current industry practices before formalizing new regulations. Consumer advocates express concern over the lack of specific anti-discrimination measures in the NAIC’s current AI bulletin, highlighting limited consumer protections regarding AI-driven underwriting and claims decisions. Meanwhile, the National Council of Insurance Legislators (NCOIL) has introduced a model act mandating human oversight over AI-related claim denials, requiring detailed documentation by qualified professionals.

The NCOIL proposal faces criticism from industry groups warning against a uniform regulatory approach that overlooks AI model diversity and the complexity of different insurance lines.

Further discussions on AI regulation in insurance are expected at the upcoming NCOIL annual meeting, reflecting ongoing efforts to balance innovation, regulatory oversight, and consumer protection in this evolving sector.