The Adjuster Was an Algorithm: The Policyholder's Bad-Faith Case Against AI Claim Handling

On-Demand · 2 CLE credits · November 12, 2026, 12:00 PM ET

An insurer cannot outsource its duty of good faith to a machine. Plead algorithmic claim handling as unreasonable conduct, trace the AI decision chain to the claim outcome, and compel the scores, audit trails, and override logs behind it.

The Claim File Is No Longer Only What the Adjuster Saw

The old question was what the adjuster was thinking. Insurers now run automated triage, severity scoring, settlement recommendations, and claim denials. The Oklahoma AG alleges that AI systems were used to undervalue claims and raise closure rates. The new question is what the algorithm was designed to do.

Biased inputs are a biased investigation. Automated undervaluation is a settlement untethered from actual damages. Optimization goals that reward lower payouts are designed misconduct, not adjuster error. Rubber-stamping the model is a failure to supervise it. Each theory tests unreasonable claim handling, inadequate investigation, and reckless disregard of the insured's interests.

You leave with four pleading theories and a causation chain that connects the technology to the claim. You leave with discovery requests built by category, deposition topics by witness, and answers to the black-box defense. Judgment calls, not doctrine.

Key topics

Agenda

Faculty

State accreditation