Exhibit A
The first exhibit of the NAIC AI Systems Evaluation Tool: a count of how many AI systems an insurer runs in each operational area, and what they are used for.
Exhibit A of the NAIC AI Systems Evaluation Tool is titled “Quantify Regulated Entity’s Use of AI Systems.” It is a counting exercise, not a governance questionnaire. It works through the operational or program areas one at a time (marketing, premium quotes and discounts, underwriting and eligibility, ratemaking and rate classification, and so on), asking for four numbers plus a description:
- How many AI system models are currently in use.
- How many have direct consumer impact, meaning they augment or automate a decision about a consumer.
- How many have material financial impact.
- How many were implemented in the past 12 months.
- The use cases those systems serve.
The tool suggests regulators begin with Exhibit A, though it says so softly: they “may wish to” use it first and then judge whether any further inquiry is warranted. Based on what comes back, they may ask for governance detail (Exhibit B), high-risk model detail (Exhibit C), or data detail (Exhibit D) where there is risk of an adverse consumer outcome or material adverse financial impact.
The practical consequence is that Exhibit A is answered from an inventory rather than being one. Nothing in the exhibit asks for a system-by-system register with named owners and validation dates. But you cannot produce honest counts without one, and the NAIC Model Bulletin asks for inventories and descriptions of the models in its own right. See our NAIC AI Evaluation Tool guide for how all four exhibits fit together.