The U.S. National Association of Insurance Commissioners (NAIC) is piloting its AI Systems Evaluation Tool.
The tool was produced by the Big Data and Artificial Intelligence Working Group, as a structured framework for insurance regulators to examine how insurers deploy AI and whether governance practices are effective.
It is a brief document, packed with specific questions across four core sections:
Exhibit A. AI Usage Inventory. A map of every AI system used: how many, in which functions, affecting which decisions. This includes vendor-embedded models and ML features inside larger platforms.
Exhibit B. Governance Risk Assessment. Frameworks and oversight structures, policies, specifics about how processes and procedures address things like residual risk, training, etc.
Exhibit C. High-Risk AI System Details. Includes claims decisions, underwriting, pricing, and fraud detection (among others). Asks about purpose and use case, who developed the model (internal or vendor), risks, testing details, validation and review, etc.
Exhibit D. AI Data Details. Provides a predefined list of data element types, and asks for information for each (type of system, description of how used, data source).
Practical and informative. Well worth a read.
Disclaimer: The info in this article is not legal advice. It may not be relevant to your circumstances. It was written for specific contexts within banks and insurers, may not apply to other contexts, and may not be relevant to other types of organisations.