Will your next claim be denied by a computer? Regulators are finally asking
Archive story from the 2026-07-16 edition. This is the reporting as filed; source publication and event dates may differ. Check the dated storyline for subsequent developments.
The NAIC's AI model bulletin has now been adopted or mirrored by 25+ states, requiring AI-supported claim decisions to comply with unfair-claims-settlement standards. California's SB 1120 bars denials made solely by automated tools without licensed human review. The policyholder bar is openly framing algorithmic denial as a bad-faith question, and regulators are probing whether human oversight is "meaningful or theater."
Why it matters
Your scope is increasingly being judged by a model before a human sees it. But the same rules that constrain carrier AI create a documented, appealable paper trail you can attack.
Our assessment
A model bulletin in twenty-five states sounds like paperwork with no teeth, and on its own it is. It also hands you a question the carrier now has to answer. Meaningful human review is the whole fight. On the next denial, ask in writing who the licensed reviewer was and what they reviewed, and put the answer, or the silence, in the file.
Saved stories and followed topics stay in this browser. No account sync or email alerts.
Legal / Regulator High confidence at publication
How we got here
Storyline: AI in claims 5 stories so far
What changed moved from the Texas DOI bulletin (6/19) to 25+ state adoption and an explicit bad-faith framing.
- When AI guesses wrong on the scope, the floor pays
- Carriers are pointing AI at your roof
- Texas regulators just put AI claims-handling on the table
- Sedgwick: 2026 CAT season will be harder to model
- Will your next claim be denied by a computer? Regulators are finally asking You are here