Once a cornerstone of financial safety and legacy planning, life insurance now faces declining relevance among consumers under 40, who will shape the industry's future. This generation isn't looking for traditional "death insurance," instead seeking solutions that deliver tangible, accessible value throughout their lives.

The data show a stark change, where over the last 15 years, life insurance's slice of individual investment wallets has dropped 23%, while equities gained 31%. The driver isn't market volatility – it's a profound disconnect between how insurers design products and how the younger demographic lives.

New research from Capgemini and LIMRA found that 68% of consumers under 40 recognize life insurance as essential for their financial future, but adoption remains stubbornly low. The reason? Traditional life insurance triggers no longer reflect the realities of next-generation policyholders: 63% have no immediate marriage plans, and 84% aren't planning children soon.

Meanwhile, only 33% of insurers recognize the growing competition from investment platforms, digital banks, and wellness subscriptions – alternatives that are rapidly gaining ground by offering immediate value, flexibility, and seamless user experiences.

This gap between consumer expectations and insurer offerings presents a significant opportunity for industry transformation.

The Living Benefits Opportunity

The disconnect is simple: Consumers under 40 want financial tools that deliver value throughout their lives, but what they get from traditional life insurance is value only at death. The solution lies in offering "living benefits" that reframe insurance as a tool for living well, not just dying well.

Younger consumers want products designed for their generation, not their parents' playbook: Cash withdrawals for life events (48%), health and wellness benefits (41%), and critical illness coverage (39%) are main priorities. They want insurance that helps start businesses, pursue wellness goals, or access fertility treatments.

Despite 76% of younger consumers expressing interest in these benefits, offerings remain limited, as the industry continues to treat them as secondary features rather than central value propositions.

Early adopters are proving that the transition to living benefits works. One European insurer redesigned income protection around "employability" rather than "unemployability," focusing on helping people stay productive rather than just covering absences. The result: two months of sales exceeded their previous full-year results. Similarly, wellness programs integrated into living benefits lead to 87% of users reporting improved wellbeing while reducing insurers' loss ratios.

Making Living Benefits Work

The need is clear: living benefits must become part of the core value proposition alongside protection. But delivering them effectively requires three fundamental changes in how insurers develop and distribute their solutions.

1. Flexible, modular solutions are the foundation of this transformation. Modern policyholders need financial solutions that evolve with their changing circumstances, adjusting when they marry later in life or when faced with unexpected career transitions. Insurers should evolve core systems around modularity and simplify underwriting to allow customers to activate, modify, or expand living benefits as their needs change, without requiring full policy rewrites or lengthy approval processes.
2. Enhanced advisory capabilities bridge the gap between product flexibility and customer engagement. While 67% of under-40s want digital access with dedicated advisor support, only 16% of insurers are equipped with these capabilities at scale. This generation wants advisors who can demonstrate how living benefits support their specific goals, which requires updating both technology platforms and compensation models to reward continuing engagement over one-off sales.
3. Ecosystem partnerships extend insurance reach beyond traditional boundaries. Rather than competing solely with other insurers, successful firms embed their offerings into the financial services, wellness, and employee platforms where under-40s already manage their lives. These partnerships transform living benefits from standalone products into integrated lifestyle tools.

The connection between these capabilities is essential. Modular products without advisory support confuse customers. Advisory capabilities without ecosystem partnerships limit reach. Inflexible products disappoint customers when they can't access promised benefits. Addressing these needs together ensures a comprehensive approach that matches how under-40s live and work.

The Moment for Action

Millennials and Gen Z are set to inherit trillions over the next decade, and many already view life insurance as a viable destination for those assets. But they won't engage with products built on outdated assumptions. They expect solutions that reflect their lifestyles, priorities, and financial goals.

Insurers face a crucial moment: Those that embed living benefits, embrace modular design, and build partnerships across wellness and financial ecosystems will earn the trust of the next generation. Those that don't, risk losing relevance to competitors that already deliver quick, lifestyle-aligned value.

The question is no longer whether life insurance matters – it's whether insurers can make it relevant to the people who will define its future.

Headlines about used car loan companies imploding may not get much attention at insurance companies, but this “old news” from the Fed should: Those companies have been making risk assessments based on wonky data and are paying the price. (The Effects of Credit Score Migration on Subprime Auto Loan and Credit Card Delinquencies). 

There is a punch line here for auto insurers. You are relying on the same wonky data, so your risk scores will likely perform worse than historically expected, too.

False negatives

A false negative occurs when a fact you intend to observe is not visible.  A classic in the literature is a pregnant woman who takes a pregnancy test that returns a negative result. In that case, you can blame the test. But a more subtle case is what the Fed is showing now, where the problem is with the data. 

Data is missing that typically weighs down a credit score and is thus driving scores higher—while the riskiness remains the same.   

Imagine a historically stable data process where good and bad observed data drive positive and negative features that calibrate a risk score. If a time or place existed where bad things were not tracked as usual (thanks, COVID) or penalties were simply less enforced (thanks, COVID), then risk scores would rise for no good reason. The COVID timeframe encouraged a period of financial transaction forbearance unlike any we have experienced in modern times.

Auto insurance has other false negatives, too. Having less enforcement of traffic rules (thanks, COVID) and less availability of traffic courts (thanks, COVID) caused similar problems with reporting on motor vehicles. For example, running red lights may have produced no tickets -- still very risky behavior, just with no typical negative indication on record. The same with speeding tickets, which haven’t been issued as frequently in recent years.

The simple equation of score = intercept + good factors - bad factors means that the absence of a bad factor mathematically leaves a score in better shape than it deserves.

Decades of observable data have been used to establish that a credit-based risk score can be useful in describing a risk scenario where the higher the score the lower the risk in an auto insurance relationship. We tend to pull credit data on assessing a new risk in policy acquisition and on a routine basis when we re-score entire portfolios of policies.

But the sort of financial forbearance that is showing up in bad loans for used cars means that people have had virtual clemency. This let many lower-quality risks appear with higher credit-based risk scores, so more risk decisions were made at terms and conditions unwarranted by the true riskiness. 

Another risk – and a new source of data

Credit data is observed backward but applied forward. A forward-looking data stream that has recently been introduced can complement existing credit-based risk assessment methods (JSI CDPI).

The stream, which grew out of work to assess the risks of students seeking loans, tracks risks based on macroeconomic effects on occupational categories. As cars were taking off a century ago, being a buggy whip maker put you at risk. Restaurant workers had a rough time during COVID. Generative AI is currently a threat to clerical workers. And so on.

Incorporating indices linked to wages and wage opportunities can help adjust the false negatives in current credit-based scores. 

For all the insurance decisions linked to credit-based scores, this may be a learning moment. 

Paul Carroll

Amid all the confusion, what can you tell us about the economic outlook and its implications for the insurance industry as we move from Q3 to Q4? 

Michel Léonard

Normally, as we wrap up Q3, we have enough data as economists, policymakers, and business leaders to start thinking about what the year will look like by the end of it. Data sometimes comes in slowly, so for the first half of the year we’re really still forecasting. But by August, normally, we’re 80%, 85%, 95% certain about where we're going to end the year. Well, that's not the case right now.

Traditionally, GDP data takes one to two quarters to firm up, but GDP doesn't move much from one quarter to the next, because an economy is a supertanker. It's not going to move rapidly. But we have tariffs this year. So we have unknowns that we haven't experienced in a developed economy like the U.S. in a long time. As a result, we expect the data to move greatly as we go into Q4. Not only the data for Q4 itself but revisions to the data for Q3 and what we already have. That really complicates things.

We went into 2025 with a shift in economic policy that had no precedent in the last 30 or 40 years. Standard economic theory, which we have to use as a framework, was telling us we could have an economic contraction on the scale of what happened during COVID. But that hasn’t happened. I don't want to tempt fate or the economic gods by saying this, but I think the worst-case scenarios will be avoided. We will end the year with a resilient U.S. economy, both in terms of growth and inflation. 

Paul Carroll

What are the key economic factors we should be looking at? 

Michel Léonard

There's a triad of unknowns: one is about GDP, one is about inflation, and one is about the Fed. And, of course, they interact with one another.

The first one really has to do with data, and, as I said, we're kind of flying blind about GDP at the moment.

The second one is data again, but for inflation. This is different. This is about inventories being depleted and how long it will take for prices to get affected and for consumer confidence to shift. We saw reports earlier this year about cratering consumer confidence, but consumers are still spending because prices haven't been affected yet. That's the second unknown. How much will tariffs increase prices? How much will be absorbed in margins? What's the elasticity of the consumer—to be a bit technical—in terms of whether they're going to keep buying?

That brings us to the third unknown in this triad, which has to do with monetary policy. The Fed is very much waiting for clarity on GDP and inflation to decide whether they should be more concerned with the risk to growth or the risk to price stability. We've been waiting for a Fed rate cut for a long time. [Editor's note: This interview occurred before the recent quarter-point cut in interest rates.] The Fed is in control of the timeline, but, at the end of the day, I think if the Fed does not cut, it's not going to be a good thing. I've been saying this for, I think, a year and a half now. 

Paul Carroll

What implications are you seeing for the insurance industry, particularly regarding replacement costs and demand? 

Michel Léonard

Traditionally, P&C lags the economy. We are slower to go into a downward cycle, and we're slower to get out of it. Right now, P&C underlying growth is outperforming overall GDP, and it's forecast to do so through next year and into 2027. But the outlook is a bit confusing. We know we're heading into a contraction of some sort. Whether that just means slower growth or means a contraction in GDP, that's unknown. The question comes down to: Is the data off, or are people actually being more resilient here in the face of tariffs and other economic issues? I'd say it's probably about the data, and we're probably going to see those forecasts for 2026 and 2027 change significantly and not for the better. [Editor’s note: This interview occurred on Sept. 5, four days before the Bureau of Labor Statistics said the U.S. economy created 911,000 fewer jobs than previously thought in the year that ended in March 2025, meaning growth was only about half as robust as previously believed.]

Paul Carroll

How is tariff policy affecting replacement costs in insurance, particularly for materials like lumber for homes and steel for automotive parts? 

Michel Léonard

On tariffs, there is no clarity whatsoever. That's part of the goal of this performative policy. We have to try to deduce the effects without necessarily passing judgment. We should continue to see a significant weakening in our expectations for homeowners insurance and for auto. On the homeowners side, prices for construction materials have increased in 2025, for the first time in four years. At the same time, folks aren't buying or expanding or renovating. With personal auto, we’ve had a great year. American consumers knew there would probably be trade tariff uncertainty and bought cars ahead of time. That buying put auto in a great place, but we probably borrowed growth in Q1 of this year and even Q2. That likely means less growth in the second half of the year and certainly next year. 

Paul Carroll

Does it even make sense to try to make a longer-term projection about the effects of the tariffs? 

Michel Léonard

It's not my job to read the president’s mind, but I do think the administration is very clear that it intends to continue using trade as tools for what it believes is best for American workers and the U.S. economy in the longer run. All economists agree that these changes will take three, four, five, six, 10 years. In the meanwhile, no matter whether the tariffs worsen or improve—by which I mean whether the tariffs get higher or lower and whether more countries are involved—they're going to remain. And once inventories from before the tariffs are depleted, we'll feel the full impact.

Paul Carroll

I’m skeptical that much manufacturing will return to the U.S. because of tariffs. Factories take years to plan and construct. An aluminum smelter requires nearly a decade. In the meantime, nobody knows whether Trump’s unilateral setting of tariffs will last.

The Supreme Court is hearing a case in November on whether Trump has unconstitutionally usurped authority that belongs to Congress; as of now, courts have ruled that Trump doesn’t have that power. Even if he escapes that threat, he faces the mid-term elections next year. His tariffs are very unpopular, meaning that Republican congressional representatives facing any kind of threat to their seat may feel pressure to reassert their authority on tariffs and force Trump to back off. And the next presidential election is two years beyond the mid-terms. Trump has promulgated all his tariffs via executive order, so the next president could undo all of them on Day One.

How many companies will build factories that are meant to last decades when Trump’s tariffs face so many challenges just in the next few years? Not many, I don’t think. 

Michel Léonard 

Reshoring was already underway before the Trump administration. It's a result of the decrease in affordable labor globally, along with rising production costs and shipping expenses. The cost difference between producing abroad versus domestically was reaching equilibrium. Now, reshoring is being accelerated by uncertainty regarding trade agreements. But people don't want the same old factory jobs. They want better jobs as a result of new factories, which will take five to 10 years to establish. And not everyone can wait five to 10 years to get to a better economic place.

Our members at the Triple-I often ask if we’ve reached a tipping point that changes the world order. Well, that's a big philosophical question. But we’ve certainly passed some milestones.

Paul Carroll

I got us off on a bit of a tangent. Any final words about the general outlook for the economy and insurance?

Michel Léonard

Just that we're going to end the year most likely in a better place than we expected, and we should be very happy about that.

Paul Carroll

Thanks, as always, Michel.

Commercial properties carry an array of risks: structural failures, natural disasters, liability claims, tenant defaults, and shifting market conditions. The margin between profitability and financial loss often comes down to how effectively risks are identified, prioritized, and mitigated. The key question is how quickly you can translate risk signals into execution.

Where Commercial Real Estate Risk Management Breaks Down

Organizations consistently face the same challenges:

• Fragmented data - Building systems, tenant records, and claims histories often sit in silos. No one has the full picture.
• Reactive approaches - Too many organizations still wait for losses to occur before acting.
• One-size-fits-all strategies - Generic risk frameworks ignore the unique exposures of high-value commercial portfolios.

These breakdowns create a gap between what risk managers know on paper and what actually protects capital. When the gap widens, losses multiply.

Execution-First Risk Management for Your Real Estate Investment

Effective strategies do not start with more technology or more checklists. They start with execution that actually works:

• Clear ownership of risks at the operational level.
• Data that feeds decisions in real time, not after quarterly reviews.
• Partnerships that understand both insurance and property operations, not just one side of the equation.

Execution-first means risk management is not a binder on a shelf. It is processes embedded into daily operations, monitored, tested, and adapted.

Core Risk Management Strategies to Protect Your Capital

● Property-Specific Risk Assessments

Standard models are not enough. Every building has unique exposures: roof condition, HVAC performance, fire suppression, and occupancy patterns. Assessments that go beyond regulatory minimums help you see where risks threaten net operating income (NOI) the most and where you should invest in mitigation efforts. Clear execution, ownership, data, and insurance reduce losses and strengthen your position as an investor in commercial properties.

Outcome: Targeted investment in mitigation where it matters most.

● Data-Driven Predictive Modeling

Claims history shows where losses have already happened, not where they will occur. Predictive models powered by IoT data, weather analytics, and building sensors let you spot emerging risks earlier. This approach helps you address issues like water intrusion or system failures before they escalate into major claims. It also ties operational data more directly to financial exposure, reserve planning, and market risk trends that can influence portfolio performance.

Outcome: Reduced surprises and better capital allocation.

● Integrated Business Continuity Planning

Risk events damage property and interrupt cash flow. Continuity plans must be tied to debt service coverage ratios, interest rate exposure, and lease obligations to reveal the full impact. Asking what happens to debt service, leases, and cash flow if a building is down for six months makes the stakes clear. Embedding continuity into contracts with vendors, lenders, and tenants helps stabilize obligations even when property damage disrupts operations.

Outcome: Faster recovery, less erosion of enterprise value.

● Tenant Risk Screening and Management

Properties are only as stable as their tenants. Credit strength, industry and market trends, and operational practices all influence property risk. Screening tenants before lease agreements helps limit exposure to weak credit and unstable businesses. Monitoring then allows you to anticipate changes early and reduce the financial risk of a market downturn or sudden rental income volatility hitting portfolio performance.

Outcome: Stronger portfolio resilience and reduced volatility in rental income.

● Insurance Optimization

Conventional property insurance programs often fail to reflect the complexity of diverse portfolios. Structuring coverage with captives, parametric triggers, and layered policies creates both flexibility and protection. A well-designed program not only manages premium-to-risk ratios but also ensures claims are paid quickly, providing liquidity at the exact moment it is needed to maintain cash flow and asset valuation.

Outcome: Balanced cost control and guaranteed capital protection.

What Leaders Should Be Asking

The best risk strategies as part of your commercial property management succeed not because of the tools but because of the people and processes around them. You should be challenging your teams and partners with questions like:

• Who owns the execution of these strategies at the asset level?
• How are we turning real-time data into decisions, not reports?
• Are our insurance structures aligned with our actual risk profile, or just market convention?
• Which potential risks could destroy the property value fastest, and how are we mitigating them today?

Protecting What Matters - Your Investment

The need to manage risk in commercial properties is not a compliance exercise. It is a capital protection strategy. When you focus on execution-first delivery, clear ownership, timely data, and strong insurance structures, you prevent losses from becoming financial shocks. Following the best practices to mitigate risk gives you an edge in commercial real estate investment, where disciplined execution separates growth opportunities from setbacks.

AI now sits at the center of the insurance conversation. Algorithms screen submissions, models calculate probability distributions, and natural language systems draft clauses and summaries. For some, this marks the beginning of a future in which machines displace the human broker, underwriter, or claims handler. That reading, however, misjudges both the nature of insurance and the limits of technology.

Insurance has never been simply about processing data. It has always been about managing uncertainty, interpreting meaning, and defending promises under pressure. AI accelerates calculation, but coherence rests on interpretation, and while algorithms may process inputs at speed, it is the broker who preserves alignment. The visible form of insurance is being reconfigured with remarkable velocity, while its strategic core remains intact.

At that core are three irreducible functions. Representation is more than the mechanical capture of data. To represent a client is to absorb their operational logic, commercial anxieties, and appetite for risk, then to embody these in a form the market can understand. No system, however advanced, can grasp what is withheld, unspoken, or only tentatively expressed. An algorithm can record information, but it cannot perceive silence. Representation is therefore an act of disciplined interpretation, giving strategic shape to a business so that it can be recognized in the language of the market.

Translation also resists automation. AI may generate summaries, but translation is not a matter of summarizing; it is about deciding what truly matters. The broker reframes exposures that are complex, partial, and ambiguous into terms the market can underwrite, and then recasts the market's response into consequences the client can act upon. This is not the simple transmission of information but the transformation of meaning. It requires knowledge of how underwriters price ambiguity, how narratives influence actuarial models, and how risk is negotiated in practice. Translation is the conversion of uncertainty into actionable form, and it cannot be mechanized.

Defense is the crucible in which insurance proves itself. When a claim is contested or a clause tested in practice, AI can accelerate document retrieval but it cannot argue meaning under pressure. A model can reference precedent, but it cannot persuade a reluctant underwriter, untangle jurisdictional complexity, or arbitrate between interpretations. Defense is existential, the moment in which alignment is reasserted precisely when promises come under strain. It is here that the difference between system and substance becomes most visible, for no technology can substitute the human broker as the final line of defense.

Representation, translation, and defense form a triad that defines the sovereignty of broking. Each carries its own risks, but only together do they hold alignment in place. If one is surrendered to automation, the others are weakened. Intake forms may simulate representation, machine outputs may mimic translation, escalation protocols may approximate defense. The outcome is substitution, not coherence.

The greater danger, then, is not that AI will disrupt insurance, but that it will induce drift. New systems accelerate transactions without recalibrating purpose. Dashboards glow green even as coverage gaps widen. Efficiency becomes a proxy for alignment. The industry risks mistaking the sophistication of AI for the substance of strategy, rehearsing coherence while quietly dissolving it.

Such drift rarely appears in the form of sudden collapse. It accumulates gradually, layer upon layer. Each unchecked assumption, each templated response, each unexamined upgrade displaces interpretive discipline a little further. The structure may remain intact, but stripped of reflection it ceases to think, functioning only in appearance rather than in substance.

The broker's task in this environment is not to resist AI but to guide its use. This responsibility involves ensuring that systems serve alignment rather than substitute for it. It requires the ability to frame, escalate, and sequence risk independently of machine logic. It also means maintaining coherence across time: managing the immediacy of renewals and claims while anchoring decisions in the slower rhythm of trust, continuity, and coverage design.

This responsibility does not mean resisting change, nor does it mean blocking progress. It is the discipline of choosing what must persist and what may be abandoned as systems evolve. It demands discernment, knowing when AI strengthens function and when it distracts from it. Sound judgment lies in the ability to hold ambiguity without rushing to premature closure, to absorb friction without reducing it to a checklist, and to defend coherence even when speed and surface efficiency suggest otherwise.

AI will remain, and its presence will expand, shaping the visible architecture of insurance with increasing force. What endures, however, is the deeper structure of the industry itself. Representation, translation, and defense cannot be automated without dissolving the very coherence they sustain. These are not optional roles; they are the functions that give insurance meaning. The firms that thrive will not be those who chase every new model first, but those who recognize what must not be surrendered at all. The future of insurance will be shaped not by the intelligence of machines, but by the discernment of those who know where technology ends and responsibility begins.

Fertilizers and pesticides have increased crop yields for centuries, but they also create broader problems that increase pollution liability risks for farmers. Legislative and legal developments in the past several decades have led to unfortunate consequences: agricultural runoff is polluting rivers and oceans, and farmers and agribusinesses are exposed to environmental liability from everyday activities that were intended only to improve food production. Retail insurance agents can play an important role in advising agribusiness clients on loss prevention practices and working with wholesale specialists to develop tailored risk transfer solutions for agriculture risks.

Risks from agricultural runoff

Irrigation and precipitation cause runoff from fields and pastures, taking on chemicals and animal waste. Eventually, this runoff finds its way into streams and rivers, and drains into oceans. The runoff can cause algal blooms that spawn low-oxygen spots and disrupt marine ecosystems.

An analysis by North Carolina State University found a complex chain of effects from the addition of nitrogen and phosphorus – common ingredients in fertilizers. Published in the Biological Review, the analysis "combined the results of 184 studies drawn from 885 individual experiments around the globe that investigated the effects of adding nitrogen and phosphorus, the main components of fertilizer, in streams and rivers. While the analysis only included studies where scientists added nitrogen and phosphorus experimentally, nitrogen and phosphorus pollution can run off from farms into streams, lakes, and rivers – as well as from wastewater discharge. At high levels, fertilizer pollution can cause harmful algal blooms and can lead to fish kills."

A prime example of these consequences is the "dead zone" in the Gulf of Mexico. The Science Education Resource Center at Carleton College says it "is primarily a result of runoff of nutrients from fertilizers and manure applied to agricultural land in the Mississippi River basin. Runoff from farms carries nutrients with the water as it drains to the Mississippi River, which ultimately flows to the Gulf of Mexico. If the number of nutrients reaching the Gulf of Mexico can be reduced, then the dead zone will begin to shrink."

The National Oceanic and Atmospheric Administration (NOAA) has studied the dead zone since 1985 and notes it varies in size. In Aug. 2024, NOAA estimated the Gulf of Mexico dead zone was more than 6,700 square miles – about the size of the state of New Jersey. Subsequently, the Hypoxia Task Force formed in 1997. Led by the U.S. Environmental Protection Agency and consisting of five federal agencies and 12 states, it has been working to implement policies and regulations to reduce the size of the zone.

Initiatives include:

• Better management of nutrient application can reduce nutrient runoff to streams.
• Planting of certain grasses, grains or clovers, called cover crops can recycle excess nutrients and reduce soil erosion, keeping nutrients out of surface waterways.
• Reducing how often fields are tilled reduces erosion and soil compaction, builds soil organic matter, and reduces runoff.
• Keeping animals and their waste out of streams, rivers, and lakes keep nitrogen and phosphorus out of the water and restores stream banks.
• Reducing nutrient loadings that drain from agricultural fields helps prevent degradation of the water in local streams and lakes.

Source: NOAA/Louisiana Universities Marine Consortium/Louisiana State University

Waste minimization and pollution risks

An analysis of United Nations' Food and Agriculture Organization data shows that global meat production quintupled from 1961 to 2023. More than 360 million tons of meat are produced worldwide from various kinds of livestock. Managing the waste produced by the livestock required to fulfill global demand is an enormous challenge.

Since 1976, when the Resource Conservation and Recovery Act (RCRA) was enacted, the U.S. has sought to address the increasing volume of municipal and industrial waste. The 1984 passage of Hazardous and Solid Waste Amendments to RCRA "required phasing out land disposal of hazardous waste, corrective action for releases and waste minimization. Waste minimization refers to the use of source reduction and/or environmentally sound recycling methods prior to treating or disposing of hazardous wastes," according to the U.S. Environmental Protection Agency.

In recent years, additional legislation has extended to the agricultural efforts of farmlands that have the continuing potential to pollute and harm the very land on which they reside, in addition to the surrounding environment, groundwater and wildlife that share these spaces. The National Resources Defense Council defines agricultural pollution as "the contamination we release into the environment as a by-product of growing and raising livestock, food crops, animal feed, and biofuel crops."

Protecting farmers' right to farm, not pollute

State and federal court decisions in agricultural pollution cases have consistently found in favor of protecting the environment and adjacent lands. Courts have generally ruled the right to farm does not grant a right to pollute. Commonly, courts levying fines have issued findings of negligence, nuisance and trespassing related to the storage, hauling, treatment and disposal of waste of all kinds.

In 2014, the Wisconsin Supreme Court ruled that "manure applied to fertilize a field in the usual course of a farming business was transformed into a 'pollutant' when it seeped into adjoining neighbors' wells. As such, the Court ruled that a 'pollution exclusion' clause in the farmer's insurance policy eliminated the insurer's duty to defend in lawsuits seeking damage for the contaminated wells."

In the underlying case, the court noted the farmer used "manure from his dairy cows as fertilizer for his fields pursuant to a nutrient management plan prepared by a certified crop agronomist and approved by the Washington County Land and Water Conservation Division. Several months later, the Wisconsin Department of Natural Resources notified the farmer that the manure had polluted a local aquifer and contaminated neighboring water wells. The well owners demanded compensation, and the farmer sought coverage from his insurer under his farm owners' policy."

The insurer sought a declaratory judgment that it had no duty to defend or indemnify the farmer because "manure was a 'pollutant' subject to exclusion under the policy. The circuit court agreed with the insurer, finding that the pollution exclusion in the policy applied to exclude coverage for damage caused by the application of manure because 'a reasonable person in the position of the (farmer) would understand cow manure to be a waste,'" the state supreme court ruled.

Mitigating pollution liability

Farmers today face many difficult issues, from the effect of climate change and weather volatility to the continued urbanization of rural areas, to how to manage runoff and waste. What can farmers do with all the manure as their operation grows? How can farmers protect themselves from pollution laws even when they've done their best to alleviate the problem?

Fortunately, tailored pollution insurance solutions are available to retail agents and brokers to help curtail the potential effect of exposures and liability confronted by their clients, today's farmers and agricultural community. Some insurance options include:

• Sudden & Accidental (Time Element) Insurance. This typically covers sudden pollution incidents that happen on an insured's land. Incidents typically need to be found in about three days and reported to the insurance carrier within two weeks. It must begin and end in a certain timeframe and is meant to exclude true gradual pollution damage claims that occur accidentally over months or years. Gradual damage tends to be the most costly, hence why Sudden & Accidental pollution coverage for agriculture risks is so inexpensive.
• Pollution Legal Liability (PLL). This is a risk management tool for property owners that is typically designed to address premises pollution exposures. This claims-made coverage, in our experience, consistently manages the on- and off-site cleanup/remediation expenses; third-party bodily injury and property damage; and defense expenses associated with industries including the agricultural sector. However, because it is an expanded gradual coverage compared to sudden and accidental, it comes with a much higher price tag.
• Transportation Pollution Liability. This type of cargo pollution insurance generally covers pollution conditions caused during transportation, loading or unloading, and sometimes mis-delivery of this cargo including waste such as manure, or a product such as fertilizer.
• Non-Owned Disposal Site Liability (NODS) typically covers disposal of waste to a non-owned disposal facility. RCRA provides for joint and several liability for waste generators, meaning a farmer that uses a non-owned disposal facility can still be deemed a potentially responsible party for cleanup and remediation by the EPA. NODS coverage exists to indemnify insureds in this situation.

The key to helping protect the agricultural community from potential pollution conditions starts with the careful understanding of the client's workplace, practices and processes. A one-size-fits-all solution may not take into account the specifics involved in an insured's work and daily practices. Other solutions include the tailoring and blending of various policy forms to create an a la carte coverage form that is designed to help cover the specific pollution conditions of this unique marketplace.

Perhaps the most convoluted conversation I've ever had occurred when my wife and I visited a tiny winery in Tuscany in the late 1990s. The winery was tucked into the stone city wall of Montepulciano, on a street barely wide enough for pedestrians, let alone our rental car. There was no sign above the tiny door, just the street number a sommelier had written down for us. An elderly man answered the door... but spoke no English, while we spoke no Italian. 

We experimented and found he understood Spanish, which both my wife and I spoke because we had recently lived in Mexico. In addition, he spoke French, which I still mostly understood from three years in Brussels. So Kim and I would ask a question in Spanish, which he translated in his head into Italian. He responded in French, which, after some fumbling, I'd translate into English. And away we went — English to Spanish to Italian to French and back to English. We learned he was a sixth-generation wine maker, heard about and tasted his wines and purchased two dozen wonderful bottles.

An announcement last week from Apple will let us use AI to cut right to the chase: translating from any language to any other language in real time and via voice, not just text. The new AirPods won't guarantee you the wonderful morning that Kim and I spent with the charming Italian winemaker but will help the insurance industry with customer service. 

The headline of the New York Times review of the Apple announcement pretty much says it all: "The New AirPods Can Translate Languages in Your Ears. This Is Profound."

The reviewer describes the AirPods as "the strongest example I had seen of AI technology working in a seamless, practical way that could be beneficial for lots of people. Children of immigrants who prefer to speak their native tongue may have an easier time communicating. Travelers visiting foreign countries may better understand cabdrivers, hotel staff and airline employees."

While the reviewer focuses on general use, not insurance, it's easy to see how the AirPods could help agents, brokers, customer service representatives and claims agents assist customers whose native language isn't English. And, while the natural tendency in the U.S. is to think about English-Spanish translation, the AirPods will, in time, be able to translate hundreds of languages into English or any other language. All you need is a set of AirPods for yourself and one to lend a client, and you can converse, with only about a one-second lag between the time you say something and when the other person hears the translation.

Translation apps of one sort or another have been available for a decade, but they've been clumsy. Some translated speech into text, which you had to then read or have the other person read, and you typically had to put your phone or other device in front of your mouth or of the person you were conversing with. The new AirPods have voice output, not only text, but mask ambient noise so well that they can be used as part of a normal conversation, some feet apart or over the phone. 

Improvements from the large language models (LLMs) used in generative AI also make the translations much more accurate than they have been, because LLMs can grasp the whole context of a conversation and not just translate individual words — which can lead to the sorts of mistakes all too familiar to anyone who's ever learned a new language. (I once got caught in a rainstorm in Mexico and, walking into the office dripping wet, told my assistant, "Estoy muy morado," when I meant to say, "Estoy muy mojado." Instead of saying, 'I'm really wet," I said, "I'm really purple." He laughed and laughed and laughed.) Individual words can matter a lot in insurance conversations and contracts, so doublechecking will always be required at some level, but the Times reviewer said his review of the transcript of a long conversation with a Spanish speaker found only tiny errors, such as whether a noun should have been translated as masculine or feminine. 

We've seen some duds among the bold attempts to embed AI in objects in the past — we're marking the 10-year anniversary of the cancellation of Google Glass, and the maker of the much-hyped Humane AI pin was sold for pennies on the dollar earlier this year — but, within a couple of years, the new AirPods should be a powerful tool for all kinds of individuals and businesses, including in insurance.

Estoy muy cierto.

Salud,

Pablo

Sometimes a picture says it all. The stock photo above shows 14 individuals working on the same vehicle and focusing on separate things. None of them knows everything the others have seen. None knows everything about the car. None necessarily documents anything they just did. They just complete a task — none talking to each other.

Sounds like every auto insurance company I have visited in the last 25 years — no exceptions.

Marketing, advertising, agency, distribution, acquisition, rate/quote/bind, service, claims, etc.: We all know those are about a car, just not exactly which car and what's inside it.

Back to the picture above: No one even counted #15, the person in the driver's seat.

That's what I'm talking about: We ignore our customer.

Worse, we blame them for not knowing their car, and we ask them car questions all the time.

It's not like we don't ignore everyone else in the insurance value chain, but not knowing the customer's car when we talk about their car with the customer seems like an epic ball drop.

We ask them all sorts of things about their car that we could just look up ourselves most of the time. Not because we are mean but because we lack basic information to communicate in a literate way about specific car features and values across our many insurance transactions.

How much empathy do we show when we use customers to do our legwork? Do we save them time? Do we add value by bringing knowledge they might expect from us? Do we show respect for their own lack of car literacy — they are just the driver. Do we serve up delight and satisfaction? Do each of us even know the features and values on our own cars? Maybe for a recent purchase with a window sticker handy, but that used car we bought second hand five years ago from a neighborhood lot with the yellow paint on the windshield that said "Sale $5,899" — of course we don't.

Face it: Nobody carries around their owners' manual. Even if they did, it doesn't say if the exterior paint is clear-coated metallic finished or just regular paint. Who really knows where the advanced driver assistance sensors are on a car — bumper, side mirror, main windshield, front grill, or multiple fused sensors, etc.? These are the clues needed for knowing you have a $5,000 windshield when a "rock from road" claim occurs versus a $500 version. This detail is how the windshield repair truck driver knows what to deliver, and what tips them off that you may need in-shop calibration.

We know that vehicles might have automatic emergency braking (AEB), but for the current car of the conversational moment, is it installed on this car? The answer is binary, yes or no, but our frame of reference is wonky… most cars on the road can't possibly have AEB. The average age of cars on the road is over 12 years old, and forward AEB technology was only entering the new car fleet sporadically about 10 years ago. Rearward AEB is even newer.

Even the new car on the lot for the new policy being issued today - is AEB there or not? While there are plans to make it a standard feature on all cars someday (like a backup camera since May 2018 in the U.S.), there are more unknowns than not, certainly if you ask the customer.

That's what I mean when I say the auto insurance industry is illiterate when it comes to cars.

Not about car insurance, but about the cars themselves.

We talk about average cars, typical cars, new cars, and features available on some cars — some available standard, some optionally available. Seldom do we talk about a specific car with specifics. 

This, even though we can just look it up. 

We have the data we already require in some policy system somewhere or in a quote flow or a claim. Literally, we can just look it up, or better still, pre-fill all our questions with the accurate manufacturer answers. So can everyone.

We struggle with "as built" and "your vehicle," but every single policy for each specific vehicle is identified by its VIN (vehicle identification number), a unique serial number assigned by the manufacturer and labeled in many ways and places on modern vehicles, including registration.

For industry veterans, there was a time long ago when there was no standard VIN (before 1981). And for many of us raised without smartphones (first iPhone was 2007) or perhaps even before the internet (World Wide Web publicly available in 1993), there are decades of model years of vehicles without electronic data that describes them deeply. But for the last decade for sure, seeing a VIN on the internet with price, features, dealer location, and window sticker details is how we shopped for our vehicle, and how we compare features and values every day.

It's time we bring to work what we have in our pockets… easy, accurate, specific details about any VIN on the street, especially the last 100 million vehicles made and sold in the U.S.

I pick on these 100 million as the NEW FLEET: the ones with all the gadgets, sensors, and a variety of possible power trains (old-fashioned internal combustion engines, mixed breed hybrids, and full-on electric vehicles). These 100 million are all model years since 2019, when being online was table stakes for car sales dealers, and a new listing had VIN and window sticker as minimal evidence that a new car was built in a certain way with a certain color and certain features for a certain advertised manufacturer suggested retail price (a base MSRP) with complete disclosure of any destination and delivery fees as well as any optional equipment installed costs to give a total MSRP.

Sure, in the real world, any manufacturer can change their base MSRP, destination and delivery fees, and option prices at any time — that they do. These updates flow onto the dealer lots and websites just like they have gone to dealerships since the very first window sticker back in the 1950s. The car sold yesterday had its window sticker, and the changes that apply today can usually be relied on to be on today's window sticker.

That sticker for a new vehicle is still there when that new car is driven off the lot on the way to its new home by its new driver. For "new fleet" used cars sold, the internet listings of its first listing have that same data. In fact, all new listings for the 100 million "new fleet" vehicles are living in a repository you can see from your smartphone right now. Most also have full "as built" manufacturers build sheet data already appended.

Show a little empathy, save a lot of time, save a lot of money. Talk isn't cheap. Neither is rework.

Let your staff and customers use the data on hand to have specific, accurate, and literate car conversations. This same data can be used to shop for replacements "just like your car" with literally no questions required to be asked. Not for shoppers, customers, agents, underwriters, actuaries, customer service people, co-workers, claims adjustors, repairers, salvage, recovery, vendors, banks, or lenders, just to name a few.

Moving from the general idea of a car to the specific "as built" details of your car is the current and future of what car literacy should look like.

As we live in an era of technological boom, it is increasingly difficult to separate hype from reality. With so many breakthroughs in medicine, space exploration and even autonomous driving, it is not advisable to bet against even the boldest of efforts. So, perhaps humans will indeed live on Mars one day. In recent times, the naysaying against hype is generally about the when it will happen, less about the if it will happen.

The P&C insurance space has seen its share of hyped concepts like blockchain and virtual reality but none as much as the recent exuberance for AI. And for good reason. Insurance practices are largely about information gathering and validation, whether for underwriting, claims or risk management. Actuarial mathematical sciences are applied for trending and pricing. Internal functions and external processes are just a few ways to describe insurance at-a-glance. All of which may benefit from AI tools and agents near-term and into the future. Insurance talent shortages, high costs of insurance for consumers and businesses, changing risks, demand for loss prediction and prevention are just some of the bigger challenges in which AI may come to the rescue.

The Gartner Hype Cycle

Most of us are familiar with the Gartner Hype Cycle, a visual model that illustrates the maturity, adoption, and social application of a new technology, charting its progression through five key phases. It provides a framework to guide technology investments by showing when a technology's actual value becomes clearer, helping organizations reduce risks and make informed decisions about when to adopt emerging technologies. As valuable as it has been, the arrival of AI technologies may challenge its relevancy.

References to AI and related application are impossible to avoid, which could lead many of us to conclude that we have moved from the Peak of Inflated Expectations and are approaching the dreaded Trough of Disillusionment in the Gartner Hype Cycle (see below).

Indeed, edge AI and generative AI have just barely entered this phase in Gartner's 2025 report.

However, it is our contention that many other uses of AI – including AI agents and decision intelligence, in insurance and beyond – could defy the history of new technologies and collapse the model, skipping right over or compressing the trough and moving straight on to the Slope of Enlightenment and the Plateau of Productivity.

This is not to say that AI adoption – we could call it commercialization at scale – is not without plenty of headwinds. Regulators and lawmakers are expressing ethical and data privacy concerns. Labor unions, service, and information workers are concerned about employment security. Insurance carrier adoption for all new technology is often quite slow, and AI is demonstrating the highest scrutiny ever. Data privacy, legal exposure and brand protection are front of mind among insurers. As legitimate as these concerns may be, collectively, they underscore the huge potential for disruption that these technologies represent.

The insurance industry is among the earliest adopters of AI, although the use cases are still somewhat basic and not yet delivering material returns. But we caution carriers not to allow these early experiences to discourage greater investment and research – the potential returns cannot be ignored, and moreover neither can the competitive market advantages.

International Insurance Society Report Identifies the Rapid Rise of AI

The International Insurance Society, which is affiliated with The Institutes, collaborated with The Institutes along with several affiliates, including the Insurance Information Institute, Insurance Thought Leadership, and Pacific Insurance Conference, on this survey.

According to their just released report, in 2025, artificial intelligence (AI) has emerged as the single most important priority among industry executives, surpassing inflation for the first time in recent years. Two-thirds of executives now place AI at the top of their technology and innovation agendas, representing a steady climb from just 17% in 2021. This accelerated focus is driven by the growing realization that AI can streamline operations, enhance data analytics, and open avenues for product innovation, all of which are seen as critical for staying competitive in an evolving business landscape.

One executive described the benefits of AI to their bottom line: "These tools enhance forecasting capabilities by allowing for deeper insights into trends and potential future risks. By empowering themselves with robust analytics, organizations can improve their strategic planning and risk management efforts, ultimately driving better business outcomes."

P&C Observations on AI Adoption

The following is a snapshot of what we are seeing and hearing from carriers and others:

• Insurers are embracing the concepts of AI to benefit in several areas, including risk selection, underwriting, operational efficiency, and cost management. Claims and underwriting tend to be the most often cited insurance use areas

• C-suites are promoting and setting mandates to advance AI agendas with a "let's not get left behind" mantra

• The variety and number of use cases for nearly every insurance function from product development to distribution and service are remarkable and optimistic

• Internal insurance carrier governance panels tend to narrow, halt and perhaps appropriately stall expansion of use cases due to data security and privacy, legal, and reputational harm risks, along with anticipated future regulatory controls

• AI for insurance encompasses a wide range of types and usage, including computer vision, generative, conversational (chatbots), agentic, and predictive

• Document review and summarization is a particular and popular use, e.g., medical records, demand packages, and legal documents

• Other areas of use include risk section, CAT modeling, claim case escalation, visual damage evaluation tools, reserving, insurance sales, and numerous internal processing functions

• Carriers are mixed in terms of buy vs. build, with most doing both by partnering with AI solutions firms and building proprietary solutions

• There is an abundance of "AI solution" providers with .ai in their URL or otherwise in marketing material, making it extremely difficult to distinguish

• Insurers are highly protective when contracting with vendors and providers, focused on data security and privacy and buckle down on AI usage, data access and related items

• Demands on people, including AI knowledge, skills and capabilities, are understated when it comes to change management

Regulators and AI

The National Association of Insurance Commissioners (NAIC) has introduced model guidance and regulations for the use of artificial intelligence systems by insurers, which have been adopted by 24 states (see NAIC Model and NAIC Model Adoption Map).

The April 2025 Genpact Consumer AI Study reveals that  a majority (55%) of U.S. adult respondents feel neutral about their insurance companies using AI, and 25% view it negatively. However, when AI delivers tangible benefits – such as faster and more accurate claims processing, customized quotes, and improved customer service – customer acceptance increases significantly. The findings emphasize an opportunity for insurers to shift perception and build preference and trust with their customers.

"As insurers embrace AI to enhance operations or customer experience, they must ensure that every interaction – whether human-led or AI-powered – meets or exceeds customer expectations," said Adil Ilyas, global business leader for insurance at Genpact. "This research highlights AI's potential to transform insurance, but also the need for insurers to close experience gaps and communicate transparently to build trust and loyalty."

Recommendations for P&C Industry Leaders

Prioritize governance and model risk management now. Regulators and plaintiffs are focused here — being proactive protects both customers and the balance sheet.

Focus first on high-value, low-risk gen-AI deployments (internal productivity, document summarization, FNOL assistance) while building the data and MLOps backbone.

Treat vendors and foundation models as concentration risk — strengthen contractual, privacy and incident response clauses.

Measure outcomes, not just outputs — track bias metrics, appeal reversal rates, customer satisfaction and financial key performance indicators (KPIs)

Plan for regulatory change — assume more granular supervisory questions and state/federal enforcement in the short term.

Disclosure and Transparency

We employed ChatGPT to gather some of the facts for this article, which itself is validation of our premise that AI is becoming pervasive and is too valuable to ignore for certain tasks. We believe that any published work should include an AI disclosure statement. When used in conjunction with the author's own experience, informed insights, common sense, and ethical judgement, it feels like AI could help make it an exciting future.

Even though insurance AI hype exceeds measurable uplift at the moment, it is our view that AI is not just another technology bubble. AI in insurance holds tremendous potential to ultimately solve the many worsening gaps and challenges in the insurance model.

Waiting for regulatory clarity or delaying to completely de-risk AI may prove to be a detrimental path with first movers having an insurmountable advantage.

Flooding is a hot topic at Climate Week NYC, as the risk of major flood events grows more frequent and destructive in the U.S., with insurance being a prevailing issue and leaders under increasing pressure to shift from reactive disaster response to proactive resilience.

In this article, I explore how U.S. risk professionals, as well as other leaders at Climate Week NYC and beyond, can learn from the U.K. in building both short- and long-term resilience.

High flood risk – low insurance uptake

Flood insurance in the U.S. is primarily offered through the National Flood Insurance Program (NFIP), run by FEMA, with some support from a small but growing private market. Homeowners, renters, and businesses in more than 20,000 participating communities can purchase coverage, whether they live in high- or low-risk flood zones.

For those in federally designated high-risk areas with a government-backed mortgage, flood insurance is often mandatory. However, uptake remains low in many other regions where flooding is still a real threat.

The majority of businesses (56%, according to a survey by Chubb) do not buy flood insurance, as they assume it is included in their commercial property policy. Many households also mistakenly believe their standard homeowners insurance covers flood damage, leaving them financially vulnerable after disasters.

Proven solutions from the U.K.

At Previsico, we have seen how the increase in flood risk has been driven by the complex dynamics of climate change, rapid urbanization, and aging infrastructure, and we are keen to share the solutions that have proved most effective in the U.K.

In the U.K., which has spent decades navigating flood risk in densely populated, flood-prone areas, we now have a proven model for integrated flood resilience with a multi-layered approach that brings together technology, tools, and stakeholders.

By sharing insights, we have been able to reduce damage, protect communities, and lower economic losses. We are also able to unlock faster, fairer recovery, and provide incentives for smarter risk management, with actionable insights.

Building flood risk into decision making

The U.K. now has both national and local policies that require developers to assess flood risk before construction, especially in flood-prone areas. Where development does proceed, it must meet strict resilience criteria, such as raised floor levels, permeable surfaces, and built-in flood defenses.

U.S. cities can benefit from this approach by embedding flood risk awareness into zoning laws, building codes, and design standards. While some U.S. cities are already doing this, the U.S. lacks a coordinated federal strategy, unlike the U.K., resulting in a more fragmented approach.

Avoiding development in high-risk areas is critical, but so is preparing new infrastructure for a wetter future. This includes climate-forward planning that accounts for future flood risk, not just historical patterns.

Enhanced building codes, requiring features like elevated electrical systems, water-resistant materials, and flood barriers, can dramatically reduce damage. Meanwhile, integrating green infrastructure, such as swales, rain gardens, and green roofs, helps to manage stormwater and ease pressure on city drainage networks.

Technology is changing the game

Early warning systems have emerged as one of the most important innovations in managing flood risk across the U.K. Forecasting is critical, particularly for storm water flooding, which is both the most common and the most difficult to predict. These systems can offer warnings up to 48 hours in advance, allowing time for vital preparations and risk mitigation. 

Accurate flood forecasting relies on highly detailed, real-time data. The most effective solutions integrate high-resolution weather forecasts, detailed topographic mapping, and advanced hydrodynamic modeling to simulate flood scenarios and predict potential impacts.

This level of precision enables targeted alerts to be sent to residents, businesses, first responders, and city officials, giving communities time to move vehicles, safeguard property, or evacuate. Lack of warning was a major issue in the recent Texas floods, which, according to AccuWeather, led to an estimated $18-22 billion of losses, capturing both direct and indirect losses, very little of which was insured.

Speed and accuracy are critical

For insurers, access to precise, real-time flood forecasting enables the delivery of alerts to policyholders, helping reduce damage and, in some cases, prevent claims altogether.

This level of accuracy is especially powerful when paired with parametric insurance. This is where location-specific data triggers automated payouts with no lengthy loss assessments. Funds provide immediate support when it matters most, such as for relocating equipment, installing flood barriers, or coordinating community evacuations.

In contrast, many people in the U.S. still rely on systems that provide alerts only after water levels rise or drainage systems are overwhelmed. By adopting predictive, hyper-local forecasting technologies, like those used in the U.K., and integrating them with parametric insurance models, U.S. cities can move from crisis response to proactive risk management.

Collaboration makes it possible

Flood resilience in the U.K. is underpinned by cross-sector collaboration, where national agencies, local governments, insurers, water companies, researchers, and community groups all play a role in managing flood risk. This holistic model results in smarter, more coordinated flood planning and response.

The U.S. can replicate this by forging partnerships with tech firms, insurers, and grassroots organizations. For example, urban planners can collaborate with climate scientists to model future risks, while insurers can provide incentives or solutions for flood-resilience and preparedness.

In the U.K., insurers are increasingly embedded in the resilience conversation, supporting initiatives like risk-reduction incentives. For U.S. insurance markets, this kind of collaboration presents a path to shared accountability, stronger risk models, and lower payouts.

A way to fast-track U.S. flood resilience

Flood resilience isn't just about building higher walls, it's about predicting risk, planning smarter, and acting earlier. U.S. cities, many of which are facing new or worsening flood threats--including New York, which has suffered greatly in recent years--can benefit hugely from this model of integrated, technology-driven resilience.

By investing in real-time flood forecasting, embedding flood risk in planning and construction, and promoting multi-stakeholder collaboration, the U.S. can not only better protect communities but also foster greater trust among governments, insurers, and the public.