Insurer Bad Faith and AI Bias: Legal Strategies to Challenge or Defend Insurer AI Tools

This session explores how insurers deploy AI technologies including machine learning, NLP, computer vision, and predictive analytics across underwriting, claims processing, and fraud detection. Attendees will learn foundational AI concepts like generative pre-trained transformers, fine-tuning, and how large language models are being adapted for insurance-specific applications.

This session examines the hallucination problem where AI fabricates information, and explores retrieval augmented generation as a partial solution. Participants will understand fundamental reasoning limitations in current AI systems and why even advanced models fail at basic tasks humans perform easily.

Learn how AI perpetuates historical biases embedded in training data, functioning as a ‘super spreader’ of past discriminatory patterns. This session covers legal implications under state anti-discrimination laws and approaches including bias audits and human-in-the-loop verification.

This session addresses the fundamental tension between insurer profit motives and good faith obligations when AI replaces human claims managers. Attendees will learn about NAIC-identified risks including inaccurate data, unfair discrimination, and lack of transparency, plus applicable legal frameworks like Unfair Claims Settlement Acts.

Examine landmark cases including Colossus, Xactimate/State Farm, and Farmers Insurance where AI-driven claims handling resulted in liability and regulatory action. These case studies demonstrate what happens when insurers rely on AI without human oversight or transparency.

This session provides practical strategies for uncovering AI use during claims processes and litigation, including pre-litigation letters and responses to confidentiality objections. Learn what to demand in discovery including validation studies, data quality documentation, and loss ratio comparisons before and after AI implementation.

Explore the risks policyholders face from their own AI use, including lawyer liability for AI errors and D&O exposure from AI disclosure failures. This session covers the Open Door case demonstrating misrepresentation liability and the governance implications of enterprise AI adoption.

A short intermission allowing attendees to refresh before the second half of the program. Use this time to review notes and prepare questions for upcoming sessions.

Learn strategies for ensuring adequate coverage by surveying AI use across business lines and working with brokers on coverage extensions. This session identifies potentially applicable traditional policies including cyber, professional liability, D&O, and media liability that may respond to AI risks.

Examine the ‘silent AI’ phenomenon and how insurers are responding with exclusions, sublimits, and AI-specific products. Learn about emerging offerings including Munich Re’s AI Sure warranties coverage and first-party model drift coverage, plus Coalition’s expanded security failure definitions.

This session provides actionable guidance for policyholders, insurers, and counsel navigating AI risks. Key takeaways include conducting formal AI surveys, over-disclosing risks in applications, and ensuring AI use complies with NAIC guiding principles.

Learn critical considerations for AI vendor contracts including data ownership, licensing rights, and liability allocation. This session addresses indemnification provisions, service level agreements, and determining responsibility when AI-related claims arise.

Explore the ongoing assessment requirements as AI models change, including periodic bias testing and remediation procedures. This session addresses privacy compliance considerations when providing data to AI systems and consumer rights to inspect, delete, and update their data.

This session covers explainable AI requirements and why avoiding black box decision-making is legally necessary for insurers. Learn how to build documentation requirements into AI platforms and ensure transparency for both regulatory compliance and customer recourse.