Overall: Swiss Re puts the global natural catastrophe protection gap at $424 billion for 2025, the widest on record.

Personal Lines: New York enacted auto reforms that ban ZIP code rating and cap injury suits to lower the nation’s highest premiums.

Commercial Lines: A London court left Chubb and Fidelis to absorb $340 million in Russia aircraft war risk payouts.

Cyber Insurance: Companies running autonomous AI agents hold a coverage class standard cyber and technology errors and omissions policies never wrote for.

The widening absolute gap alongside an improving resilience ratio may complicate how chief risk officers and investor relations teams frame catastrophe exposure to boards, since the headline number rises every year even when the company is covering more of the risk than before.

Product and innovation leaders may find the gap concentrated in high growth, catastrophe exposed states is precisely where admitted capacity is thinnest, leaving parametric and specialty teams to build in markets their own carrier is non renewing.

Reinsurance and treaty buyers may face a harder pricing argument, because a gap driven by asset values rather than under insurance weakens the case for blanket rate relief at renewal.

The record secondary peril share could shift internal authority toward catastrophe modeling and exposure management functions, whose loss estimates now carry more weight than the historical loss triangles actuaries have leaned on.

Distribution leaders in exposed geographies may see agents reframe the gap as a sales opening, which can pull marginal risk into the book at the point where reinsurance attachment points are least forgiving.

For carriers weighing public private partnership structures, the data may strengthen regulators who argue private capacity alone will not close the gap, narrowing the room to decline politically sensitive risks without a state backstop conversation.

Digital Now Half of New Policy Purchases J.D. Power released its 2026 U.S. Insurance Digital Experience Study on May 13, 2026, finding that 47 percent of new auto and home insurance policies are now purchased through digital channels, ahead of agents at 35 percent and call centers at 17 percent. Overall satisfaction in the service segment fell 4 points to 695 on a 1,000 point scale. Amica ranked highest in service at 730. The study, based on 11,553 evaluations fielded from January through March 2026, found that only 11 percent of customers used chatbots or virtual assistants, though those who did reported higher satisfaction.

The prohibition on ZIP code, occupation, education, and homeownership as primary rating variables may force pricing actuaries to rebuild segmentation around behavioral and vehicle data, a multi quarter systems project that competitors with cleaner data pipelines could finish faster.

Underwriting and product leaders may find the severity reforms and the rating factor ban pulling in opposite directions, with one improving loss costs while the other compresses the pricing precision that made the New York book selectable.

The requirement to explain any increase above 10 percent may hand competitors a recurring opening, since every large increase now arrives with a documented rationale a rival agent can rebut.

General counsels and compliance leaders may carry the heaviest near term load, because the staged crash and excess profit provisions create new enforcement surfaces that expose carriers to penalties for practices that were previously routine.

For carriers that priced New York as a managed loss leader inside a multistate book, the reforms may shift the internal economics enough that capital allocation committees revisit how much growth the state should carry.

Distribution leaders may see projected savings used as an acquisition message before any rate moves, creating an expectation gap that falls on agents to manage when filed reductions lag the messaging.

Liberty Mutual Puts Auto Quoting Inside ChatGPT Liberty Mutual Insurance announced on May 28, 2026 the launch of what it calls the first carrier backed conversational AI auto insurance quoting application inside ChatGPT. The seventh largest U.S. auto insurer said drivers can obtain a quote through a chat based exchange that runs on the company's own rating engine, then finish binding at LibertyMutual.com. The app is live in seven states, Arizona, Kentucky, Ohio, Missouri, New Mexico, Utah, and Wisconsin, with plans to expand to more than 40 states by the end of 2026. Tyler Asher, chief distribution and marketing officer for U.S. Retail Markets, framed the move as extending customer access.

The ruling may push specialty underwriters and product managers to draft contribution and indemnity rights between insurers explicitly, since the court showed that paying first does not preserve a direct claim against co insurers who did not pay.

Reinsurance and treaty structures for war and confiscation risk may face renewed scrutiny from chief risk officers, because losses that courts continue to litigate years later complicate how reserves and recoverables are carried.

The forced reliance on subrogation rather than direct contribution may lengthen the tail on these claims, leaving claims and finance teams running recovery actions in the name of insureds long after the payouts cleared.

For carriers writing contingent war risk behind operator policies, the decision may change how that layered structure is priced, since the contingent layer now looks more exposed and less recoverable than the cascade assumed.

The precedent may influence how Lloyd’s syndicates and London market participants structure shared risk programs on future geopolitical exposures, where the order of response between policies now carries clearer financial consequences.

Investor relations and finance leaders at the affected insurers may face questions about how much exposure remains open, given the parallel appeal and the gap between amounts paid and amounts recoverable.

Carriers File Broader AI Exclusions Beyond ISO Forms Several carriers have filed their own artificial intelligence liability exclusions with state regulators since ISO's three generative AI endorsements, CG 40 47, CG 40 48, and CG 35 08, took effect in January 2026 across commercial general liability (CGL) policies. W.R. Berkley introduced Form PC 51380, an absolute AI exclusion across directors and officers (D&O), errors and omissions (E&O), and fiduciary liability that bars any claim arising from the use, deployment, or development of artificial intelligence by any person or entity. AIG and Great American have filed their own versions, and Berkshire Hathaway, Chubb, and Travelers have secured state approval to strip AI related damages from standard commercial policies. The Intelligence Council reported on ISO's near universal AI exclusion adoption on May 18, 2026

The absence of standard wording may concentrate pricing power with the few underwriters who publish coherent agentic frameworks first, letting them set reference terms the rest of the market negotiates against before any claims history exists.

Cyber and technology E&O teams may find the silent coverage zone becoming a reserving problem before it becomes a product, since a single agent driven loss could be argued into a policy that was never priced for autonomous action.

The EU timeline may split the book by geography, leaving multinational insureds and their brokers to reconcile carriers that treat August 2026 as binding against those waiting to see whether the Digital Omnibus defers the obligations.

For chief underwriting officers, agentic AI may blur the line between cyber, technology E&O, and general liability, since one autonomous agent can produce a loss that three different policies each have reason to disclaim.

Claims executives may need autonomous system forensic capability before the actuarial data to support it arrives, because adjudicating an agent driven loss means reconstructing decisions no human directly made.

Carriers that move early on affirmative agentic coverage may attract exactly the most AI intensive insureds, a selection pattern that concentrates an untested exposure in the books of the carriers most willing to write it.