CHICAGO - Plaintiffs in a federal court in Illinois defended their motion to amend a complaint that alleges their insurer's use of artificial intelligence subjected minorities to additional scrutiny, arguing that discovery produced new data supporting the claims. The case centers on whether automated claims or underwriting processes disproportionately flagged policyholders based on protected characteristics.
The plaintiffs filed the motion after reviewing documents and data obtained during discovery, which they say backs up the original allegations with specific evidence. The insurer has opposed the amendment, and the court has yet to rule on whether the revised complaint will be accepted.
The dispute over AI-driven scrutiny
The original complaint accused the insurer of deploying an AI system that applied different levels of review to minority policyholders. The plaintiffs say the system generated higher rates of additional scrutiny - such as extra documentation requests, delayed claims, or more frequent audits - for those policyholders.
According to the filing, the newly produced discovery material includes internal records that show how the AI model was trained and tested. The plaintiffs argue those records reveal patterns that were not fully visible when they first filed the complaint.
The insurer has not yet filed a full response to the amended allegations. The case is one of a growing number of disputes in which courts are being asked to weigh whether an algorithm's output constitutes discriminatory conduct under existing insurance law.
What the discovery adds
Legal analysts following the case say the core issue is whether the plaintiffs can show the AI system caused the disparate treatment, rather than the system merely reflecting broader underwriting practices. The amendment appears aimed at closing that gap.
Courts have split on how to treat AI-based discrimination claims. Some have required plaintiffs to show the algorithm itself was designed with discriminatory intent. Others have accepted statistical evidence of disparate outcomes as enough to move a case forward.
The plaintiffs' filing argues that the new data shows the AI system assigned risk scores that correlated with race in ways that could not be explained by legitimate underwriting factors. The insurer's legal team has argued that the data is being taken out of context and that the algorithm is regularly audited for bias.
Why this matters for insurance professionals
For insurance professionals, the outcome of this motion will help define what regulators and courts expect from AI governance. If the amendment is allowed, insurers may need to prepare for discovery requests that dig into model training data, feature selection, and audit trails. That affects how compliance teams document AI decisions.
This case also signals that plaintiffs' attorneys are getting better at using discovery to build AI discrimination cases. Insurers should review their own models for patterns that could be interpreted as disparate impact, and document the business rationale for any factor that correlates with protected class. The question is no longer whether AI tools are fair, but how their fairness is proven in court.
For those following the legal side of these disputes, the AI for Insurance resources track similar enforcement and litigation developments. The AI for Legal section covers the broader wave of algorithmic accountability cases.
Your membership also unlocks:
