AI is too risky to insure, say people whose job is insuring risk

Major U.S. insurers including Great American, Chubb and W. R. Berkley have asked regulators for permission to exclude broad AI-related liabilities from corporate insurance policies, citing the models' opaque outputs and the risk of systemic simultaneous claims. The move follows high-profile incidents — a $110 million lawsuit over a false Google AI allegation, a $25 million CEO-cloning fraud, and a chatbot pricing error — and industry warnings that carriers can absorb single losses (Aon cited $400m) but not thousands of concurrent claims from an agentic AI failure, creating downside risk for insurers and corporate AI adoption.

Analysis

Market structure is set to bifurcate: large diversified brokers/reinsurance platforms and specialty carriers that accept AI risk will gain pricing power, while broad-market P&C writers (e.g., CB, WRB) face margin pressure as they exclude or limit AI exposures; expect 3–8% EPS downside risk for exposed underwriters over 12 months if exclusions become widespread. Demand will shift toward bespoke AI policies, captives and parametric products, tightening supply of traditional coverage and allowing specialty carriers to charge 20–40% premium surcharges for first-loss AI risk in early rollouts.

Tail risks include a systemic agentic failure or an adverse court ruling that expands insured losses—low probability but >$1bn balance-sheet hit for several carriers simultaneously—forcing capital raises or dividend cuts within 6–18 months. Near-term (days–weeks) volatility will be driven by regulatory filings and press cycles; medium-term (3–12 months) by underwriting guideline updates and reinsurance treaty rewrites; long-term (1–3 years) by legislation clarifying producer/vendor liability and indemnity practices.

Trade implications: favor brokers/reinsurers that monetize advisory and structuring (AON) and selective specialty reinsurers that can underwrite AI with granular limits; underweight legacy P&C carriers with broad commercial books (CB, WRB). Use 3–9 month options to express view: buy puts on specific insurer equities to hedge tail risk and buy call spreads on AON to capture fee upside from captive/cyber demand expansion.

Consensus misses that exclusions create a nascent multi-billion-dollar specialty market—early entrants can earn outsized ROEs; conversely, market reaction may be overdone if vendors accept stronger indemnities or reinsurers layer capacity, limiting insurer losses. Historical analogs (asbestos/environmental) show pain for decades but eventual creation of specialist carriers and securitization vehicles that can be long-term winners.

Key Decisions for Investors

• Establish a 2–3% portfolio short equity position in CB and a separate 1–2% short in WRB sized to risk tolerance; implement immediately and scale to full size within 10 trading days to capture repricing momentum from underwriting notices.
• Buy 3–6 month puts on CB and WRB ~7–12% OTM sized to cap downside to 1% of portfolio each (cost-controlled hedge against clustered claims or regulatory shock); reassess at 90 days or after any adverse court/regulatory decision.
• Initiate a 1.5–2% long position in AON (equity) and concurrently buy a 6-month call spread (AON, strike range +5%/+20%) to capture advisory/restructuring fee upside as clients seek captive/specialty products; target exit at +25% gain or after 12 months.
• Rotate 2–4% of insurance exposure into cyber/security vendors (e.g., CRWD or MSFT security unit) and specialty reinsurers with capital flexibility (e.g., RNR/RE) over next 30 days; prioritize names with proven ALM and capital buffers and trim after 6–12 months or if spread compression >100bp.
• Monitor three triggers over next 30–90 days before scaling bets: (1) regulator responses to exclusion filings, (2) two or more public class actions quantifying AI losses >$100m, (3) reinsurer treaty language changes — increase position size if reinsurers offer explicit AI capacity (reduces tail) or cut exposures if regulators ban exclusions.