Insurers, Plaintiff Bar Wage AI Arms Race

Key Takeaways

• 78% of P&C insurers already use generative AI, only 4% scaled
• LLMs ingest unstructured case files, outpacing legacy rule engines
• Plaintiff platform EvenUp valued >$2 B, crowdsources settlement data
• 25 states plus D.C. enforce NAIC AI Model Bulletin for insurers
• AI-driven settlement gaps could compress nuisance‑value claim economics

Pulse Analysis

The insurance industry is at the forefront of a generative‑AI transformation that could unlock $50‑70 billion in new revenue, according to McKinsey. While Bain reports that 78 % of property‑and‑casualty carriers have already integrated AI tools, only a handful have moved beyond pilot projects. Legacy systems such as the rule‑based Colossus engine, which still processes the majority of claims, are being eclipsed by large language models that can read entire case files—medical records, deposition transcripts, and police reports—without predefined fields. This leap in analytical depth is compressing the cost of claim valuation and reshaping the economics of civil litigation.

The traditional data moat—insurers’ exclusive access to millions of closed claims and settlement histories—is eroding as plaintiff‑side innovators crowdsource outcomes. EvenUp, now valued at more than $2 billion, aggregates settlement figures from over 2,000 firms handling roughly 10,000 cases weekly, while CLARA Analytics builds a shared database across its carrier clients. As both sides assemble comparable training sets, the informational advantage that once allowed insurers to dictate settlement ranges is narrowing, forcing negotiations to rely more on model outputs than on opaque internal benchmarks.

Regulators have responded quickly. Over 25 states and the District of Columbia have adopted the NAIC’s Model Bulletin on AI, mandating governance, consumer notice, and bias testing; Colorado’s SB 21‑169 even threatens civil penalties for non‑compliance. The emerging legal landscape will likely reward firms that can explain their model’s logic and demonstrate fairness. For litigators, the strategic imperative is clear: develop or acquire comparable AI capabilities, audit opponent models, and stay ahead of regulatory requirements, or risk ceding decisive leverage in an increasingly algorithm‑driven dispute resolution arena.