One of the major consequences of a changing climate is the damage caused by severe storms and flooding. Coastal communities can be particularly difficult for insurers to serve, given the differences in construction standards from one structure to the next.
AI is helping insurers navigate these troubled waters by taking these nuances into consideration and making sense of updated weather data. Read more below.
AI helps private insurers fill flood coverage gaps before NFIP deadline
With NFIP funding set to expire in September, private insurers are deploying AI to expand flood coverage into high-risk areas where insurance isn't mortgage-mandated — including communities like Ruidoso, N.M., and Asheville, N.C. Aon, which is investing $1 billion in technology and analytics over three years, uses AI to process hazard, exposure and loss data for granular risk accumulation views. Outdated flood maps, some more than a decade old, represent a key opportunity: AI can incorporate current weather data to improve pricing accuracy. Lilypad Insurance reports AI has compressed product development cycles from 12 months to roughly three months, a speed advantage that could prove critical as the September deadline approaches.
Read more:
AI-driven property-level pricing can stabilize coastal markets
With $137 billion in global insured catastrophe losses in 2024 and 15% to 20% of Florida homeowners going uninsured rather than absorbing inflated premiums, territory-based pricing is producing compounding market failures — not rounding errors. Florida's state insurer of last resort peaked at 1.4 million policies in 2023, a scale it was never designed to handle. The path forward requires AI-powered, property-level underwriting that processes satellite imagery, elevation data, structural attributes and live weather inputs to surface granular risk signals for underwriter review. Carriers that adopt this approach can write coastal business profitably; those that do not will continue retreating, enlarging the public backstop they were never meant to replace.
Read more:
AI cuts auto claim cycle times from five days to one or two
AI-powered estimating tools are compressing auto claim cycle times from more than five days to one or two days — a performance gap that's becoming a retention lever as premium-sensitive policyholders grow more willing to switch carriers. The tools assess photo submissions for visible damage, flag documentation gaps before estimates are finalized, and reduce costly supplements and rework. Carriers deploying AI beyond estimating are also automating invoice processing, fraud detection and proactive status notifications. The most effective implementations pair AI with human adjusters rather than replacing them.
Read more:
83% of breaches involve AI; carriers lag on their own cyber controls
AI is implicated in 83% of reported security breaches, with breach rates up 40% over three years, yet only 30% of affected organizations had adequate response capabilities, according to Gigamon research. For carriers, brokers and TPAs — which collectively hold the most monetizable concentration of financial, medical and identity data in any sector — the exposure is acute. Fifty-six percent of insurance companies had at least one compromised credential in the past two years. Genuine readiness in 2026 requires credential lifecycle controls, documented third-party access audits, verifiable enforcement of attested controls, and formal access revocation processes. Carriers pricing a projected $19.6 billion global cyber market in 2026 must apply that same underwriting rigor internally.
Read more:
29% of firms hit by AI-assisted cyberattacks, QBE finds
Nearly a third of U.S. businesses experienced an AI-assisted cyber incident in the past year, according to a QBE survey of 400 SMB leaders — with the finance-and-insurance sector accounting for 8% of commercial espionage incidents between 2021 and 2026. The threat vector is agentic AI, which unlike generative AI can act autonomously, run simultaneous campaigns across platforms and create new attack vectors without requiring technical expertise from bad actors. QBE recommends a layered "back-to-basics" security posture centered on identity access management, behavioral monitoring and AI-enabled threat detection, paired with comprehensive cyber coverage to close exposure gaps.
Read more:
92% of health insurers use or plan to use AI, but workflows come first
AI adoption in insurance is nearly universal — 92% of health insurers, 88% of auto insurers and 70% of home insurers report using, planning to use or exploring AI or machine learning, per the NAIC. But adoption without operational readiness produces faster errors, not better outcomes. Carriers with fragmented systems across claims, billing, policy administration and onboarding should prioritize connecting that infrastructure before layering in AI tools. The NAIC also makes clear that insurers remain legally responsible for regulatory compliance and consumer protection when using AI — meaning governance, transparency and defined human oversight aren't optional. Clean data and accountable workflows determine whether AI delivers value or accelerates existing problems.
Read more:
Allianz Partners to cut up to 1,800 jobs amid AI push
Allianz Partners, the assistance and travel insurance subsidiary of Allianz SE, plans to eliminate 1,500 to 1,800 positions across Europe through severance agreements, early retirements and voluntary departures. CEO Tomas Kunzmann cited AI implementation as the primary driver, affecting employees in Spain, France, Germany, Italy and the Benelux countries. The unit employs more than 22,000 people. The announcement follows a similar move by Munich Re's Ergo unit, which is cutting roughly 1,000 German positions partly due to AI adoption. Bloomberg Economics estimates 27% of workers in advanced economies face meaningful AI-related displacement — a signal for insurance executives to assess workforce restructuring timelines and labor relations strategies now.
Read more:
This roundup was created with AI assistance. A Digital Insurance editor reviewed each item before publication.
