Regulation — model law wave
NCOIL AI Systems in Insurance Model Act — the next state-legislation wave
The National Council of Insurance Legislators adopted its AI Systems in Insurance Model Act in 2025. Where the NAIC Bulletin is a supervisory expectation, an NCOIL model act is model text state legislatures actually enact. Several states are already carrying enabling bills.
What the model act actually says
"An insurer that uses an AI system in an insurance practice shall establish, implement, and maintain a written AI systems program that addresses governance, risk management, and internal controls …"
NCOIL AI Systems in Insurance Model Act, Section 4 (adopted 2025) — program mandate
"The insurer shall conduct testing, on a periodic basis, of AI systems used in insurance practices to assess the potential for unfair discrimination or bias … and shall address any such outcomes identified through the testing."
NCOIL AI Systems Model Act, Section 5 — testing + remediation
"The Commissioner may examine an insurer's AI systems program to determine compliance with this Act … The insurer shall make available to the Commissioner all records, documentation, and other information necessary for the examination."
NCOIL AI Systems Model Act, Section 7 — examination authority
What this means in plain English
The model act pulls the NAIC Bulletin's supervisory expectations into hard state law:
- Written AI program. Statutorily required, not just supervisorily expected.
- Periodic testing for unfair discrimination or bias. Explicit cadence + remediation.
- Commissioner examination authority. The state DOI can inspect the AI program the same way it inspects rate filings.
Because it is model text, the language will show up nearly identically in whichever state enacts. That is the point of a model act.
What triggers the exposure in the sample
I3 group decline / refer-SIU rate rose 11% to 62% silently over 60+ days. Once the NCOIL model act is enacted in any state the carrier writes in, three separate examination items are now live: is there a written AI program? has periodic testing been done? are the results + remediation documented? The carrier's answer needs to be "yes, here" — not "we're working on it."
What the $499 Snapshot shows against this rule
- Independent-verifier record — the "periodic testing" evidence the model act requires
- Written thresholds + methodology + results + steps taken — the audit trail the Commissioner examination will look for
- Signed independent-verifier declaration — the third-party statement that makes the record inspection-ready
- 3 fix-first items scoped to the AI surface — the remediation starting point
See the 5-decision reproducibility drill →
$499 Snapshot. 3 business days.
Independent-verifier determination scoped to your carrier's AI surface + written record ready for Commissioner examination.
Buy $499
Snapshot credit applies to Baseline ($2,500) or Enterprise Attestation ($35-55K) upgrade within 30 days.