Court Allows Discovery Into Insurer’s Use of AI to Deny Claims

Why It Matters

The order gives policyholders a concrete mechanism to challenge AI‑driven claim denials and exposes insurers to potential bad‑faith claims. It also forces the industry to adopt transparent AI governance to avoid costly litigation and regulatory exposure.

Key Takeaways

• Federal court permits discovery of insurer AI usage
• AI program nH Predict documents partially disclosed
• Courts view AI chat files as non‑privileged
• Policyholders can challenge AI‑driven claim denials
• Insurers must safeguard privileged data from AI platforms

Pulse Analysis

The Minnesota federal court’s ruling in *Estate of Gene B. Lokken v. UnitedHealth Group* marks a watershed moment for the insurance sector’s adoption of artificial‑intelligence tools. By ordering the production of documents that explain how the proprietary nH Predict system evaluates post‑acute care claims, the court signaled that AI‑driven decision‑making is no longer a black box shielded from scrutiny. The decision aligns with a recent New York precedent that classified AI chat logs as discoverable, reinforcing a judicial trend that treats algorithmic outputs like any other evidence in coverage disputes.

For litigators, the ruling reshapes discovery strategy. Plaintiffs can now compel insurers to disclose AI development policies, training materials, and oversight mechanisms, while defendants must be prepared to justify the accuracy and fairness of algorithmic outputs. The court’s emphasis on whether AI “supplants physician decision‑making” suggests that any lack of human review could be framed as bad‑faith conduct, especially when an AI “hallucination” leads to an erroneous denial. Companies should therefore audit what information they feed into AI platforms, as privileged data entered into non‑confidential systems may be waived.

The broader market feels the ripple effect. Insurers that rely heavily on machine‑learning models must now balance efficiency gains against the risk of costly discovery and potential regulatory scrutiny. Transparent AI governance—clear usage policies, documented human oversight, and secure data handling—will become a competitive differentiator. Meanwhile, policyholders gain a powerful tool to challenge opaque denial decisions, potentially shifting the balance of power in coverage litigation. As courts continue to demystify algorithmic decision‑making, the industry can expect tighter standards for AI accountability and increased investment in explainable‑AI technologies.