The Lawletter Vol. 51 No. 2
Amy Gore—Senior Attorney
With the corporate push to assimilate artificial intelligence (“AI”) into business, the insurance industry has started embracing the use of various AI tools in underwriting and certain claim-handling applications. However, traditional liabilities continue to be a risk to insurance companies.
In Estate of Lokken v. UnitedHealth Group, Inc., 766 F. Supp. 3d 835, 840 (D. Minn. 2025), the claimant alleged that United Health Group, Inc.’s use of AI to deny an insurance claim under a Medicare Advantage policy amounted to a breach of the insurer’s implied covenant of good faith and fair dealing. There, the claimant asserted that the insurance company used a program called “nH Predict” to deny a claim rather than have the submitted documentation reviewed by a physician, and that such reliance on the AI program amounted to common law bad faith. A similar claim was asserted against Humana in Estate of Barrows v. Humana, Inc., Civil Action No. 3:23-cv-654-RGJ, 2025 U.S. Dist. LEXIS 158565, at *1 (W.D. Ky. Aug. 14, 2025). In both cases, the court ruled on procedural grounds without addressing the bad faith accusations, although in Estate of Lokken, the claimants were permitted to proceed with their breach of contract and breach of the implied covenant of good faith and fair dealing claims.
These actions highlight a new area in the field of insurance bad faith litigation. Has the insurer relied on an AI-based program in its underwriting or claims resolution process? Furthermore, if an AI program was used by the insurance company, will it be insulated from a potential bad faith claim? What new discovery requests will be needed? Whether by statutory or common law standards, an insured seeking to recover under a bad faith claim must generally prove that the insurer's conduct was unreasonable, frivolous, or unfounded, and that the breach of the duty owed to its insured caused damages. See generally Fla. Stat. § 624.155; State Farm Mut. Auto. Ins. Co. v. Floyd, 235 Va. 136, 137, 366 S.E.2d 93, 93 (1988). While relying on the use of AI-generated programs may streamline the insurer’s claims-handling process, strict reliance on such programs can open the insurer up to further litigation in much the same manner as attorneys who carelessly rely on AI programs for their briefs. Ford v. Sherwin-Williams, No. 25-1022-DDC-GEB, 2026 U.S. Dist. LEXIS 69427, at *1 (D. Kan. Mar. 31, 2026); Hill v. Auto Club Fam. Ins. Co., No. 2:24-cv-00107-KS-BWR, 2025 U.S. Dist. LEXIS 182282, at *18 (S.D. Miss Sept. 17, 2025); Malone-Bey v. Lauderdale Cnty. Sch. Bd., No. 3:25-cv-380-KHJ-MTP, 2025 U.S. Dist. LEXIS 142684, 2025 WL 2098352, at *3 (S.D. Miss. July 25, 2025).
