North Dakota Makes 29: The NAIC AI Bulletin and What It Means for Insurance Agencies
Prices and facts checked October 13, 2026. We don't accept payment for rankings.
On this page
What happened: on October 5, 2026, North Dakota adopted the National Association of Insurance Commissioners' Model Bulletin on the Use of Artificial Intelligence Systems by Insurers, following South Carolina (September 24), Mississippi (July 22) and Puerto Rico (June 11). The NAIC's adoption map, updated October 8, 2026, now counts 29 jurisdictions that have adopted the bulletin, plus 4 states with their own insurance-specific AI rules (NAIC adoption map).
The NAIC adopted the bulletin on December 4, 2023 (NAIC). Adoption has been steady: 24 states had signed on by early 2025 (Quarles), and 25 states plus D.C. by March 2026 (Lumenova).
Why it matters to agencies: the bulletin is aimed at insurers, but it makes carriers responsible for AI supplied by third parties. As agencies add AI receptionists, quoting assistants and AI-written client emails, carriers will increasingly ask what you use, and regulators can ask carriers for that documentation.
Not legal advice. Adoption and wording vary by state, and some large states run their own AI frameworks. Check your state's insurance department and your carrier agreements.
What the bulletin asks of insurers
| Area | Expectation |
|---|---|
| Written AI program | A documented program for responsible AI use, covering governance, risk management and internal audit, proportionate to how much AI is used and the potential consumer harm |
| Governance | Clear accountability, with senior management answerable to the board and cross-functional oversight |
| Risk controls | Data quality, bias, testing and validation, monitoring for model drift, privacy and record retention |
| Consumer notice | Telling consumers when AI is used, in a way suited to the stage of the process |
| Third parties | Due diligence on AI vendors, contract terms for audit rights and regulatory cooperation, and actually using those rights |
| Exams | Being able to produce the AI program, validation records and vendor files when regulators ask |
Where agencies come in
The bulletin doesn't regulate producers directly. But two parts reach you in practice:
- Vendor oversight flows down. Carriers stay responsible for AI used in regulated insurance decisions, including vendor-supplied AI. If your AI tools touch quoting, underwriting information or claims intake for a carrier, expect questions or contract terms (Quarles).
- Consumer notice. If AI is part of the customer's experience, such as an AI voice answering calls and collecting quote details, being upfront about it fits the bulletin's direction and protects you.
Four large states use their own insurance-specific AI rules instead: California (Bulletin 2022-5), Colorado (3 CCR 702-10), New York (Circular Letter No. 7) and Texas (Bulletin B-0036-20) (NAIC adoption map).
Who has adopted it (as of October 8, 2026)
| Adopted the bulletin | Own insurance-specific AI rules |
|---|---|
| Alaska, Arkansas, Connecticut, Delaware, D.C., Hawaii, Illinois, Iowa, Kentucky, Maryland, Massachusetts, Michigan, Mississippi, Nebraska, Nevada, New Hampshire, New Jersey, North Carolina, North Dakota, Oklahoma, Pennsylvania, Puerto Rico, Rhode Island, South Carolina, Vermont, Virginia, Washington, West Virginia, Wisconsin | California, Colorado, New York, Texas |
Source: NAIC adoption map. If your state isn't listed, your carriers may still apply the bulletin's standards across all the states they write in.
What regulators are building next
The bulletin is only part of the NAIC's AI work. Its AI Systems Evaluation Tool, which helps regulators collect information on insurers' AI use, governance, high-risk models and data inputs, was being piloted by 12 states as of March 2026, with adoption expected at the 2026 Fall National Meeting. A separate working group is developing a framework for regulators to evaluate third-party AI data and models (NAIC).
For agencies, that points one way: more structured questions about AI, reaching further down the chain to vendors and distribution partners.
An agency checklist for AI tools
- Inventory every AI tool: receptionists, chat, quoting helpers, email drafting, CRM automation.
- Classify the data each one touches: personal information, quote details, claims information.
- Ask each carrier whether it has approved-tool lists or AI requirements for agencies.
- Get vendor documentation: security report, data-use and training terms, retention, breach notice.
- Disclose AI to customers where they interact with it, for example "this call is answered by an AI assistant."
- Keep humans on decisions: coverage advice, binding and claims judgments stay with licensed staff.
Outbound AI calling has separate federal rules: the FCC has ruled that AI-generated voices count as "artificial" voices under the TCPA (Hunton). See AI voice calls and the TCPA. Choosing tools? Compare AI receptionists for agencies.
FAQ
Does the NAIC bulletin apply to my agency?
It's directed at insurers. Agencies are affected indirectly, through carrier vendor oversight and contracts, and through any state-specific rules.
Is the bulletin law?
It's model guidance that states adopt as bulletins or regulations, sometimes with changes. Where adopted, regulators expect insurers to follow it.
How many states have adopted it?
29 jurisdictions (27 states, D.C. and Puerto Rico) as of the NAIC's October 8, 2026 map. Check the NAIC map for the current list, since states keep adding it.
Sources
Pages we used for this article, checked on or before the update date above.
- NAIC adoption map · content.naic.org
- NAIC · content.naic.org
- Quarles · quarles.com
- Lumenova · lumenova.ai
- Hunton · hunton.com
Changelog: Oct 13, 2026: first published.