AI is rewriting what’s possible in insurance, from smarter workflows to more intuitive customer experiences. Here are summaries of the latest news and insights to help keep industry leaders in touch with the next wave of intelligent innovation.
News: Could the “Las Vegas Model” drive the cost of stand-alone insurance coverage out of reach?
The root of it: Embedded insurance is gaining momentum, and it could pressure traditional insurers if manufacturers and tech giants bundle coverage at steep discounts – or even offer it as a “loss-leader” to sell more of their core products. Using the “Las Vegas Model” (giving away one product to sell more of another), brands like Apple, Tesla, and Amazon can treat insurance as a cost to eliminate, leveraging superior data and direct customer relationships to possibly undercut stand-alone carriers over the next 5–20+ years.
News: Insured NatCat losses top $100B for 6th straight year
The root of it: In 2025, insured natural catastrophe losses totaled $107 billion, marking the sixth consecutive year that losses exceeded $100 billion. Losses were driven primarily by record-setting Los Angeles wildfires and severe convective storms, despite hurricane activity remaining unusually low. Although total losses were 24% lower than in 2024, the continued streak of high-loss years highlights the sustained financial impact of natural catastrophes on the insurance industry.
News: Rate pressure is driving lower-income homeowners away from NFIP
The root of it: A new study finds FEMA’s Risk Rating 2.0 flood pricing reforms, which aim to align National Flood Insurance Program (NFIP) premiums with actual flood risk, have led to significant policy drop-outs – especially among lower-income homeowners facing steep increases. Up to 13% of those hit hardest by rate hikes are abandoning coverage, potentially widening already significant coverage gaps in flood-prone areas.
The root of it: As AI adoption accelerates, regulators are approaching similar AI risks – bias, accountability, and consumer harm – via wide-ranging solutions. The EU is enforcing a binding, risk-tiered AI Act, while the UK favors sector-led oversight within existing frameworks. The US regulatory environment is pitting state-level rules against the sweeping authority of the federal government. And Singapore is advancing supervisory, capability-driven guidance for financial institutions. Combined, these disparate methods could increase global insurers' governance overhead burden.
News: “Dark speculation” introduces an insurance-inspired framework for confronting frontier AI risk
The root of it: Harvard professor Daniel Carpenter proposes “dark speculation,” a framework for assessing extreme AI risks that avoids both hype and panic. Borrowing from insurance underwriting principles and practices, it combines scenario modeling with probabilistic reasoning and explicit assumptions about institutional response. By emphasizing mitigation, data gaps, and likelihood – and not just worst cases – the approach reframes frontier AI risks as severe but potentially manageable, supporting clearer policy, investment, and governance decisions.
News: A closer look at New York and California's new “AI companion” laws
The root of it: New York and California are the first states to regulate AI companions – chatbots that simulate ongoing, personal interactions. The laws require clear AI-identity disclosures and crisis safeguards, with California adding youth protections, reporting requirements, and private rights of action. As a recent federal AI executive order raises questions about long-term preemption, individual states’ rules are nevertheless moving ahead to set near-term compliance expectations for AI-driven engagement tools.
News: Are there valid arguments for chatbots’ free speech “rights”?
Few would dispute whether AI chatbots can generate false or offensive speech – they clearly can and often do. At issue is whether output from AI models should be treated differently from that of humans who knowingly generate falsehoods constantly without prior restraint. This article argues that recent political reactions to chatbot misstatements reflect discomfort with scale and automation, but not with a novel category of harm. The risk of action against automation (rather than “intent”) is that informal government intervention can quietly expand into broader censorship standards without transparency or due process.
News: Is third-party litigation funding turning US courts into a capital market?
The root of it: An opinion piece appearing in The Hill argues that third-party litigation funding is reshaping US courts by attracting hedge funds, sovereign wealth funds, and foreign investors who finance and influence lawsuits for profit. The authors cite the rapid growth in litigation finance, its role in mass torts, limited transparency, and concerns that outside capital – including foreign sources – may distort settlements, judicial independence, and access to justice.
News: Insurance innovation is more about culture than tech
The root of it: Insurance innovation suffers less from immature technology and more from cultural entrenchment. Carriers often hinder progress through risk-averse mindsets, siloed teams, “not-invented-here” bias, and treating innovation as a side project instead of core work. In an industry heavily dependent on technology but built around caution and stewardship, meaningful change requires operating models that provide teams with the room to test, learn, and iterate – without undermining stability.
News: These three AI trends will define insurance in the coming year
The root of it: In 2026, insurers will shift from AI pilots to enterprise scale as three trends converge: interoperability, AI-native core systems, and human feedback loops. With 67% of carriers testing AI but only ~7% scaling it, open architectures and standards like Model Context Protocol can plug new underwriting or claims tools into workflows without lock-in. AI-native cores enable real-time data, traceable decisions, faster product launches – while structured human oversight builds explainability, fairness, and regulatory trust.
News: McKinsey report shows how agentic AI is transforming the customer experience for a troubled industry
The root of it: This study examines how agentic AI – AI systems capable of autonomous decision-making and action – could reshape customer experience in travel, an industry in which airlines routinely face operational complexity and high rates of customer dissatisfaction. Agentic systems can manage entire functions – such as booking changes during disruptions – rather than isolated tasks. While travel companies are experimenting with AI, only a minority see a measurable financial impact so far. Clear signs of early value are emerging in customer service, sales support, and end-to-end process redesign.
News: Clarity on AI liability is testing carriers, especially in D&O and E&O
The root of it: AI-related exclusions are showing up more frequently in commercial insurance policies, but they haven’t yet had a significant impact on cyber insurance. In fact, many carriers are clarifying that cyber losses involving AI are still intended to be covered. Experts note that AI attacks, such as deepfakes and social engineering, amplify existing cyber risks. However, the bigger concern is the emergence of broad AI exclusions in D&O, E&O, management liability, and other lines. Despite this trend, industry leaders caution against panic – current cyber coverages generally still protect against traditional exposures, and the focus is on modernizing existing products rather than creating standalone AI lines.
News: Lawyers’ “productive laziness” in using AI is creating liability headaches
The root of it: Attorneys’ increasing use of AI in legal work is creating serious professional risks, mainly when lawyers rely on AI output without verifying accuracy. US courts have sanctioned lawyers for submitting AI-generated work with fabricated citations, highlighting the danger of what some call “productive laziness,” as AI shortcuts replace diligent review. (Legal malpractice insurance may not cover sanctions or claims arising from careless use of AI.) Attorneys are advised to verify all AI-generated research and bill accurately to avoid ethical and coverage issues.
News: Bain & Co. report shows P&C carriers are missing out by only “dabbling” in generative AI
The root of it: Bain’s 2025 Claims Maturity Assessment concludes that most P&C carriers are still “dabbling” in generative AI, with 78% of surveyed insurers using it "in some capacity," as opposed to just 4% who report scaling it enterprise-wide. Today’s deployments are mostly narrow (document summarization, customer communications support, fraud detection). Bain’s analysis suggests the bigger payoff comes from redesigning processes end-to-end (e.g., claims).
News: Yet another private equity firm enters the reinsurance arena
The root of it: Brookfield’s Oaktree Capital is expanding its presence into reinsurance by agreeing to invest hundreds of millions of dollars to reinsure policies sold by Allianz, via a new vehicle at Lloyd’s of London. The move underscores the growing appetite of alternative investors for reinsurance capacity, echoing a similar deal struck last year by AIG and Blackstone, as private capital seeks an expanded role in how insurers cede and finance risk.
News: Florida Gov. DeSantis pushes back on POTUS’s federal AI regulation executive order
The root of it: Gov. Ron DeSantis argued that a pending Trump executive order on AI cannot preempt state lawmaking, insisting only the US Congress has that authority. As Florida lawmakers explore new AI protections in insurance and other areas, DeSantis criticized federal attempts to block states without proposing a coherent regulatory framework. With a federal order imminent, state legislative leaders say they are pausing to assess how it could affect their authority.
News: Florida’s state legislature advances its AI claims-handling bills
The root of it: Lawmakers in Florida’s House of Representatives and state Senate are advancing two bills – HB 527 and SB 202, respectively – to set guardrails for insurers’ use of AI in insurance claims. The bills would bar algorithms, AI, or machine learning from being the sole basis for adjusting or denying a claim, requiring a “qualified human professional” to independently analyze the claim, review the accuracy of any AI output, and confirm reasons why nonpayment or reduced payment is warranted. If the bill passes, Florida's Office of Insurance Regulation (OIR) would verify insurers’ compliance.
News: California workers’ comp combined ratio hits 20-year high of 127%
The root of it: As reported by Risk Placement Services (RPS), California’s workers’ compensation combined ratio surged to 127% in 2024, the highest in more than two decades. This deterioration reflects rising medical costs, reserve strain, litigation, cumulative trauma claim severity, among other factors. Sustained pressure on carrier profitability contributed to the first state rate increase in a decade (8.7% in 2025), with carriers raising renewals and tightening underwriting as market conditions continue to firm.
News: NAIC moves forward with its 2026 AI evaluation pilot program
The root of it: State insurance regulators, through the National Association of Insurance Commissioners' (NAIC) Big Data and Artificial Intelligence Working Group, are proceeding with a 2026 pilot of an AI Systems Evaluation Tool to assess how insurers use AI. Ten carriers are expected to participate in the program, with results driving refinements to the tool, which could later be used in market conduct and financial exams. This initiative faces industry objections over transparency, structure, and potential additions to insurers' existing regulatory burden.
News: The Pacific Northwest is being pummeled by record snow and rainfall – with more on the way
The root of it: Back-to-back atmospheric rivers are forecast to drench the Pacific Northwest and northern California through December 20, bringing 4–8+ inches of rain, significant snow accumulations, and extended heavy precipitation to parts of the region already reeling from the effects of recent extreme weather events. Snowfall amounts could reach 24–48 inches at high altitudes, and models show continued risk of additional atmospheric river activity later this month.
The root of it: President Donald Trump plans to issue an executive order establishing a single national AI rule, overriding the patchwork of state laws that the tech industry argues could hinder innovation. Companies, including OpenAI, Google, and Meta, support federal standards, warning that the US risks falling behind China. However, leaders in both parties defend state authority as they advance privacy and deepfake laws. A previous attempt to curb state AI rules met overwhelming bipartisan resistance, culminating in a 99–1 Senate vote against it.
News: Insurers explore AI chatbot error liability exclusions for 2026
The root of it: Several commercial carriers are moving to limit their exposure to losses caused by AI chatbots, introducing broad policy exclusions that remove coverage for erroneous outputs such as faulty quotes or incorrect medical guidance. Measures proposed for 2026 would include exclusions for bodily injury, property damage, and advertising injury tied to generative AI systems. Yet uncertainty remains, as regulators, brokers, and courts will ultimately determine what qualifies as “AI use” and how far these exclusions can reach.
News: Cyber risk continues to stay one step ahead of cyber insurance
The root of it: Cyber risk is evolving faster than cyber insurance can respond, with specialized threat actors, supply chain vulnerabilities, AI, and other hazards exacerbating existing vulnerabilities. Sector-specific exposures in healthcare, retail, utilities, and energy are rising as once-isolated operational technology (OT) networks become more interconnected. Insurers are tightening controls and tailoring policy wording, but third-party losses, inconsistent pricing, and year-round risk management demands show coverage is still chasing the risk.
News: Fitch Ratings outlook on US P&C markets is neutral for 2026, despite robust profitability
The root of it: Fitch Ratings has maintained its “neutral” outlook on the US P&C insurance market for 2026. This assessment is based on robust 2025 results, specifically, a projected combined ratio of 93.7% (nearly a three-point improvement), healthy reserve releases, favorable personal auto performance, and a “mild hurricane season.” Looking ahead, Fitch expects a modest rise in the combined ratio to 96–97% in 2026. However, insurers may face pressure from softer pricing, weaker investment income, and macroeconomic risks such as economic slowdown, competition, and social inflation.
News: Zillow discontinues posting climate risk scores for listed properties
The root of it: Zillow has removed the on-site climate-risk scores from more than one million home listings. The scores – sourced from climate risk modeler First Street – had rated properties’ exposure to floods, wildfires, wind, heat, and other hazards. The decision follows pushback from the California Regional Multiple Listing Service (CRMLS) and real estate agents, who argued some scores appeared inaccurate and negatively affected property sales. Zillow no longer embeds the risk scores directly in listings, instead providing only a link to First Street’s site for climate-risk information.
News: Florida is urged to “stay the course” with insurance reforms, but some issues remain unresolved
The root of it: At the Florida Chamber of Commerce Insurance Summit, insurers, reinsurers, and bond market leaders urged lawmakers to preserve the 2022–23 property reforms that curtailed litigation and stabilized loss costs. The state-run insurer of last resort, Citizens Property Insurance Corporation, has shed over 1 million policies as private carriers re-enter the market. Speakers at the conference cautioned against repealing Florida’s PIP “no-fault” auto law as a 2026 bill threatens its future, potentially jeopardizing reform momentum.
News: Nvidia CEO Huang predicts 90% of the world’s knowledge could come from AI before 2030
The root of it: In a recent interview, Jenson Huang forecasted that within two or three years, around 90% of the world’s knowledge – in textbooks, reports, and other content – could be generated by AI. He argued that synthetic and human-written sources are effectively equivalent as long as the content is fact-checked and grounded in “fundamental first principles.”
News: A NY appellate court decision limits insurers’ authority to deny no-fault claims
The root of it: New York’s Court of Appeals ruled 6–1 that insurers cannot unilaterally deny no-fault claims based on alleged provider misconduct. Only state regulators, such as the Board of Regents, have the authority to make those determinations. The case involved a $390,000 kickback scheme tied to $4.9 million in acupuncture bills. Insurers can still employ remedies such as challenging medical necessity and reporting suspected provider fraud to regulators, but they must route most misconduct concerns through the state's process.
News: Embedded insurance projected to be a trillion-dollar market by 2033
The root of it: The embedded insurance market is now forecast to reach about $1.1 trillion in global gross written premiums (GWP) by 2033, or roughly 15% of the total global GWP figure. This is a dramatic upward revision from prior forecasts of approximately $700 billion in GWP by 2029, reflecting strong growth in auto insurance sold at the point of purchase. This suggests that embedded insurance is rapidly evolving into a significant growth engine for insurers, insurtechs, and other ecosystem partners.
News: Record flooding in Thailand disrupts auto parts and tech supply chains
The root of it: Thailand’s worst-ever southern floods have killed more than 180 people and crippled key logistics hubs, halting exports of auto parts and tech components. Hat Yai, a crucial commercial and transport gateway, is effectively cut off, with roads submerged and rail services suspended. Authorities warn exporters “simply cannot move their shipments,” risking lasting damage to Thailand’s role as a regional supply hub as buyers shift to suppliers based in Indonesia and Vietnam.
News: NYC Comptroller’s report highlights the hazards and impact of e-delivery services
The root of it: New York City Comptroller Brad Lander’s “Fast Shipping. Slow Justice” report links booming e-commerce and last-mile e-delivery services to rising motor vehicle incidents, worker injuries, and pollution. After these distribution centers open, vehicle-related injuries nearby rise by 16%, truck crashes increase by 146%, and warehouse incident rates exceed three times the national average. Roughly two-thirds of the facilities sit in Environmental Justice Areas, concentrating risks in already vulnerable neighborhoods.
News: What can be done about those “meticulous” AI rental car inspections?
The root of it: Hertz’s rollout of AI-powered vehicle inspections has sparked outrage, with renters reporting that they have been billed hundreds of dollars for minor or invisible "damage" that was previously considered acceptable wear and tear. The article’s author contends the problem isn’t the technology, but how it’s used. With improved transparency, human-in-the-loop reviews, and clear communication of what constitutes billable damage, this powerful tool could deliver consistent, objective vehicle condition records to benefit insurers, service shops, and future renters.
News: “The ultimate [impaired] driving machine?” – BMW owners top DUI list
The root of it: A recent study conducted by a criminal defense/personal injury law firm shows that BMW drivers receive the most US drunk-driving citations – over three per 1,000 drivers – with other high-end vehicle makes also topping the list. A spokesperson for the study’s sponsor notes that these brands appear more frequently in DUI data, aligning with a University of California study finding that drivers of luxury vehicles stop less often for pedestrians.
Read our 2025 State of AI Adoption in Insurance Report for insights and perspectives on AI adoption from more than 240 insurance executives.