New York, May 7, 2026
Market shock and a policy retreat
Insurers across the United States are quietly shrinking the reach of umbrella and excess liability lines after a wave of oversized jury awards and the industry's first wave of artificial intelligence exclusions. Carriers are raising attachment points, narrowing forms and in some cases refusing to write standalone umbrella policies for higher-risk clients, a retreat that is already reshaping program design for fleets, manufacturers and high-net-worth individuals.
Why limits are shrinking
The commercial casualty market has been under sustained pressure from social inflation and what the industry calls "nuclear verdicts" - jury awards that exceed $10 million. Independent research shows a sharp jump in these verdicts in 2024, with 135 cases totaling about $31.3 billion, the largest annual tally on record. That spike has put visible strain on reinsurers and primary carriers, prompting more conservative limit deployment and tougher underwriting on exposures with high jury risk.
At the same time, market-watchers report that capacity for standalone umbrella layers is limited and that many insurers prefer to offer modest lead limits only when underlying risk controls and loss histories are exemplary. Brokers now routinely face higher retentions and requests for layered towers rather than single-carrier cushions of high limits. Pricing remains uneven, but attachment points are moving upward across high-hazard classes.
New exclusions for AI accelerate the shift
Insurers are also responding to a fast-growing and uncertain exposure: losses tied to artificial intelligence. Standard-setting organizations and several carriers filed new policy language last year to carve out certain generative AI risks from commercial liability forms. Those endorsements - now available for insurers to adopt in many states - have the effect of shrinking the scope of coverage for incidents connected to automated models and generative outputs unless buyers purchase separate, affirmative AI coverage. Underwriters say the move is intended to force clearer pricing of novel technology risk.
That carve-out is already changing negotiations. Risk managers who had considered umbrella layers a straightforward, inexpensive way to protect corporate balance sheets are discovering carriers want explicit disclosures about AI use, model governance and third-party oversight before committing higher limits.
What buyers and brokers are doing
Brokers are rebuilding excess towers, adding quota-share layers and turning to alternative capital solutions where traditional markets are constrained. Companies with commercial fleets, heavy auto exposure or operations in litigation-prone jurisdictions are being steered toward higher self-insured retentions, more stringent loss-control programs and, in some cases, captive structures to retain more risk. Specialized AI policies and bespoke cyber-technology riders are emerging as stopgap options where exclusionary language appears.
For insureds, the consequence is concrete: higher out-of-pocket risk, more paperwork at renewals, and the need to shop programs earlier. For the market, the combination of nuclear verdicts and AI uncertainty is accelerating a broader re-pricing of liability risk and a move toward more granular, contract-specific underwriting.
Bottom line
The liability market is in transition. The twin forces of record jury awards and the industry's first-generation AI exclusions are narrowing traditional umbrellas and forcing buyers, brokers and capital providers to redesign protection strategies. The change is not yet uniform, but it is widespread enough that risk leaders and legal teams are reprioritizing program architecture for the year ahead.