Insurance companies are sharing their AI success stories with a focus on specific niches:
Companies are finding that there are some customer interactions that should still involve humans, and final underwriting decisions can't be handled by software. Additionally, a lot of AI deployments remain unfinished, leading employees to go rogue and use unsanctioned AI tools.
Read more from our coverage below.
AI retention models help insurers combat rising customer churn
As agentic AI makes it easier for policyholders to shop and switch carriers, insurers are deploying AI-driven churn, retention and next-best-action models to get ahead of defection. LexisNexis VP Brad Hames notes that previously loyal customer segments are now shopping around, raising the stakes for carriers. Chubb uses AI to segment prospects by product and geography, accelerating identification of high-value markets. Gallagher's commercial clients, meanwhile, are demanding data-backed advisory services — the firm responds with AI-powered analytics tools, including a customer portal and an industry-specific planning system that generates carrier-ready risk-improvement recommendations.
Read more:
Frontline Insurance routes 70% of FNOL calls to AI — and keeps humans for the rest
With insured catastrophe losses exceeding $100 billion for the sixth consecutive year and experienced adjusters retiring faster than replacements can be hired, the pressure to maximize automation is real — but chasing 100% autonomous completion rates risks eroding policyholder trust. Frontline Insurance offers a working model: voice AI handles 70% of FNOL calls, responding in roughly one second, while complex or emotionally charged situations transfer immediately to a human. J.D. Power's 2026 U.S. Property Claims Satisfaction Study reinforces the point — speed alone doesn't drive satisfaction. The more meaningful metric isn't automation rate; it's how quickly a distressed customer reaches the right person.
Read more:
The Hartford cuts underwriting time with AI in Q2 2026
The Hartford reported $1.3 billion in net income for Q2 2026 — up 30.5% year over year — as executives credited AI-enabled workflows with completing underwriting activities "in a fraction of the time" across middle-market and large commercial lines. The insurer declined to provide specific performance metrics but described investments as "significant." The results reinforce a broader industry pattern: Travelers developed its own in-house large language model trained on proprietary insurance documents, while UnitedHealth Group is deploying $3 billion in AI initiatives through 2027. For commercial lines leaders, the competitive pressure is clear — brokers and customers increasingly expect faster submission turnarounds without sacrificing pricing discipline.
Read more:
Only 23% of P&C insurers have fully integrated AI, report finds
Despite rising AI budgets, full integration remains rare across the P&C sector — and the gap is costly. Only 23% of insurers report fully integrated AI, while 89% of employees acknowledge using unsanctioned AI tools outside approved systems, signaling significant governance risk. Fragmented deployments are driving up costs as AI operates parallel to, rather than within, core systems. The payoff for closing that gap is measurable: Fully integrated insurers are 3.6 times more likely to achieve real-time portfolio control. Meanwhile, 31% of carriers take weeks or longer to detect shifts in portfolio performance — a lag that integrated AI can directly address, according to Federato's 2026 State of P&C Insurance Technology report.
Read more:
5 barriers blocking commercial insurers' AI transformation
Legacy infrastructure, data fragmentation, workflow integration gaps, regulatory exposure and talent shortages are the five primary obstacles stalling enterprise-wide AI adoption among commercial insurers in 2026. Moving from proof-of-concept to scaled deployment requires modernizing core systems through cloud migration and APIs, establishing rigorous data governance frameworks, and embedding AI directly into underwriting, claims and policy administration workflows. On the compliance side, explainable AI models and human oversight protocols are non-negotiable given regulators' heightened scrutiny of decision transparency and bias. Bridging the skills gap means hiring professionals fluent in both insurance operations and machine learning — and building internal cultures where AI augments, rather than threatens, human expertise.
Read more:
Guardrails, not replacement: How AI should work in underwriting
Insurance lags hedge funds and top global banks in AI adoption by several years, but that gap creates an opportunity to leapfrog legacy transformation, says Nirmitee Shah, general manager of financial and professional services at Scale AI, which counts Allianz among its clients. The highest-value targets: distribution, underwriting and claims. In underwriting specifically, the priority is defining what AI cannot infer — not just what it can do — to prevent pattern-recognition errors that create operational and regulatory risk. Shah also recommends treating AI as an end-to-end process orchestrator rather than a system replacement, and advocates for NAIC-led regulatory sandboxes to accelerate compliant adoption.
Read more:
Don't write off the metaverse: lessons from ERGO Group
Despite Meta's retreat from its original metaverse vision, ERGO Group argues the broader ecosystem — spanning extended reality, immersive platforms and AI wearables — remains a live opportunity for insurers. Roblox reports 380 million monthly active users; Fortnite, 650 million — signals that immersive digital environments are scaling regardless of any single platform's fate. ERGO has already deployed VR-based training to more than 3,000 staff since mid-2023 and launched a mixed-reality risk-education app for the Apple Vision Pro. Carriers exploring the space should look beyond headset-centric platforms toward extended reality and immersive experience — and start building internal expertise now, before market demand forces a rushed response.
Read more:
Climate peril severity hits 7-year high, spurring AI model adoption
Climate peril severity surged 26% from 2024 and 93% since 2019 — the highest in seven years — according to LexisNexis's latest home trends report. Although loss costs fell 4% and frequency fell 24% in 2025, analysts caution that trend will not hold as billion-dollar weather events multiply. Carriers seeking to control peril costs need granular, by-peril property intelligence rather than reliance on macro-level averages. AI-powered visual language models are making digital self-inspection tools increasingly viable for assessing property condition and risk-hardening measures before losses occur — offering a cost-effective alternative to in-person inspections. Machine learning is also enabling higher-resolution flood and terrain simulations, giving underwriters sharper risk differentiation at the property level.
Read more:
AI could add $2B in premiums by 2030 with connected data
Agentic AI embedded in life insurance distribution could increase new annualized individual life premiums by 11% by 2030, generating an estimated $2 billion in incremental annual premiums — but only if carriers integrate AI across connected data and end-to-end workflows. The opportunity is significant: Nearly 100 million Americans are uninsured or underinsured, and 40% overestimate the cost of a basic term policy. Fragmented planning, illustration and underwriting systems leave advisors without a complete consumer picture. Building interoperable infrastructure with consistent data governance, defined consent protocols and cross-functional alignment is the prerequisite for deploying AI that can personalize guidance and accelerate placement.
Read more:
This roundup was created with AI assistance. A Digital Insurance editor reviewed each item before publication.
