NAIC AI Evaluation Tool Pilot Expands Toward Fall Adoption

The NAIC’s Big Data and Artificial Intelligence Working Group continues to develop the AI Systems Evaluation Tool, the questionnaire examiners use to assess how insurers govern AI. It is running as a multi-state pilot, with national adoption targeted for the November 2026 NAIC fall meeting. Carriers should expect the tool’s four-exhibit structure to shape how AI examinations are scoped.

The tool is a supplemental examination framework, not a new regulation. It attaches AI-specific questions to the market conduct, financial condition, and financial analysis handbooks regulators already use. The four exhibits follow a sequence. Exhibit A asks for a quantitative inventory of AI systems. Exhibit B asks for a governance risk assessment. Exhibit C drills into high-risk systems. Exhibit D asks for data source and lineage details. Each exhibit assumes the one before it is complete and honest. If the inventory in Exhibit A is missing systems, the governance assessment in Exhibit B covers the wrong scope, and the high-risk detail in Exhibit C will be questioned.

For carriers, the practical implication is that preparation must start with the inventory, not with the questionnaire. You cannot describe governance over systems you have not counted. You cannot detail a high-risk model you have not classified as high-risk. The carriers that will struggle in the first wave of exams are not the ones with weak policies; they are the ones with weak inventories, where a system the company uses every day was never logged as AI because it was licensed from a vendor or embedded in a larger workflow.

The pilot matters because the states running it are stress-testing the questions before national adoption. Feedback from the pilot could sharpen wording, add examples, or change which documentation is treated as sufficient. Carriers that wait until November to prepare will be reacting to a finished instrument. Carriers that prepare now can influence how their own data is presented and, in some cases, identify gaps that are easier to close before the exam notice lands.

The immediate work is to walk your governance against the tool’s checklist elements and mark what you can evidence versus what you cannot. A short gap list with owners and dates is more useful than a polished narrative, because it tells the examiner you have already done the hard thinking. The longer work is to connect the four exhibits inside your own program so the inventory, governance, high-risk testing, and data lineage all tell the same story.

One practical way to start is to run a tabletop exercise. Ask the team that owns the model inventory to produce the list an examiner would see in Exhibit A. Then hand that list to the governance committee and ask them to walk through Exhibit B using only the systems on the list. If the committee cannot map its oversight to the inventory, the seam is real and needs to be closed before the exam notice arrives. The same exercise works between Exhibit C and D: pick one high-risk system and trace its data sources, validation history, and drift monitoring back to the governance minutes.

We track the working group’s materials weekly. For the full breakdown of what each exhibit asks, see our guide to the NAIC AI Evaluation Tool.