Artificial intelligence has entered a new era. It's no longer just a statistical predictor crunching historical data. It's now a creator, planner, and autonomous actor capable of generating content, making decisions, and executing multi-step tasks. This leap from traditional AI to generative and now agentic AI has fundamentally changed the risk landscape. These new AI systems therefore demand a rethink of how we measure, manage, and insure the risk.

Traditional insurance frameworks, predominantly built on backward-looking data and well-understood failure modes, are not suited for systems that learn, adapt, and change behavior in real time. As AI becomes more deeply woven into business, infrastructure, and daily life, the question is no longer if it will fail but how and who bears the cost when it does.

To unlock the full potential of AI safely and at scale, the insurance industry must innovate. This is not just about transferring financial risk, but also about creating market incentives for trustworthy AI adoption. Insurers and risk managers will need to deploy new tools to quantify, price, and monitor AI exposure, ensuring that innovation and safety evolve together. This is urgent, as many AI risks are sitting silently inside existing policies, often unpriced, unmanaged, and waiting to materialize. The systemic risk posed by this silent coverage represents a significant, largely unmodeled aggregation exposure for carriers and creates uncertainty for the insured.

In the sections that follow, we explore how the AI risk profile is evolving and why a new generation of assurance and insurance mechanisms will be critical to building confidence in the intelligent systems that will increasingly shape our world.

The Evolving AI Risk Profile

AI systems have gone through three major generations, each more capable and complex than the last. With every step, the risk profile has expanded.

1. Traditional AI: Early AI systems were essentially statistical predictors. They learned patterns from structured data to forecast outcomes -- for example, credit scores, demand forecasts, and spam detection. Their risks were relatively stable and easy to quantify, mostly limited to data quality problems or model misspecification.
2. Generative AI: Generative AI (e.g. large language or diffusion models) doesn't just analyze data; it creates content. This creative power comes with new risk: producing plausible but false outputs (hallucinations), reusing copyrighted material from training data, or shifting behavior as APIs or retrievers change over time. Because these systems are composable (built from multiple moving parts) and dynamic (updated frequently), they can change behavior without warning.
3. Agentic AI: The newest wave, agentic AI, adds autonomy, reasoning, and tool use. Autonomy brings systemic risk: small local errors can cascade across an entire chain of actions, a phenomenon known as compounding uncertainty. When such systems fail, tracing the root cause or conducting causation analysis becomes extremely difficult due to opaque failure modes and information asymmetry.

The critical challenge is that AI can now fail while doing exactly what it was designed to do. Unlike software bugs or cyberattacks, these failures emerge from within due to biased training data, drifting knowledge, or complex feedback loops. Managing such behavior requires continuous, evidence-based oversight rather than static, one-off testing.

From Checklists to Continuous Monitoring

For AI systems to be insurable or trusted in safety-critical domains, they must undergo rigorous, transparent, and repeatable AI risk management. That means moving from checklist validation to continuous monitoring, where systems are tested and challenged throughout their lifecycle. This risk management framework provides the necessary evidence and controls that underwriters will demand to price the exposure.

Best practice frameworks point to four foundations, which should be viewed as future underwriting criteria:

• Governance and Tiering: Treat the whole workflow from data pipelines to prompts and APIs as the governed unit. Tier systems not just by impact but also by autonomy (how much they act without human approval) and volatility (how often components change). Every modification should trigger a change-impact review.
• Design Standards: Start from intent: what is "failure" in business or operational terms? Translate that into measurable technical metrics, justify every heuristic (prompt templates, data filters, reward models), and document assumptions and known residual risks. Build guardrails and fallback plans from day one.
• Validation Uplift: Move beyond static benchmarks. Combine domain-grounded tests with adversarial evaluation and scenario stress-testing. Measure calibration and selective prediction; use red teaming to expose hidden vulnerabilities. Where LLMs are used as judges, demand statistical checks for bias and consistency.
• Monitoring: Deploy continuous monitoring across inputs, outputs, and dependencies. Track drift, fragility, and anomalous behavior. Establish clear service-level objectives for safety and accuracy. Keep humans in the loop for escalation and design rapid rollback and patching playbooks.

In this new landscape, model probe systems for blind spots, test procedural reliability, and pressure-test entire pipelines. The goal isn't just compliance, it's resilience: building AI systems that remain safe, and trustworthy even as they evolve. Experience in managing cyber risk means insurers can build on existing practices, but tools and methods will need to be adapted to AI systems.

The Case for AI Insurance: Turning Risk into Resilience

As AI systems become more autonomous and unpredictable, they test the limits of traditional insurance models. Losses caused by AI errors often don't fit neatly into existing policy lines like cyber, product liability, or professional indemnity, therefore creating coverage uncertainty. This often results in "silent coverage," which creates hidden liabilities, unpriced exposures, and uncertainty for both insurers and insured. This unreserved, unmodeled exposure threatens aggregation events and solvency for carriers.

From our perspective, it matters less whether AI risks eventually sit within existing policy lines, emerge as embedded features, or evolve into a new, standalone class of AI insurance. What matters is that AI risks are material and growing, creating significant exposure to portfolios and businesses alike. As such, they must be rigorously understood, quantified, and managed. Businesses adopting AI will need confidence that, when failures occur, clearly defined insurance coverage stands behind the technology if they decide to transfer the risk into the market.

To make AI risk insurable, the market will need innovative tools and pricing mechanisms that reflect how AI operates:

• Performance-Based Guarantees: Policies could trigger payouts if the AI underperforms (e.g., if its accuracy or reliability drops below a defined threshold). This mechanism could be structured as an endorsement on Product Liability or a custom Financial Loss policy.
• Usage-Based Insurance: Premiums can scale with AI activity (e.g., per API call, per decision), creating dynamic, real-time pricing that mirrors exposure levels.
• Premium Differentiation (Bonus–Malus): Safer systems should cost less to insure. Firms that can demonstrate robust governance, transparent validation, and effective monitoring would pay lower premiums. In contrast, opaque or unaudited systems would be priced prohibitively high or deemed uninsurable.

This market mechanism does something regulation alone cannot: it aligns financial incentives with technical rigor. Underwriters will demand strong assurance, continuous monitoring, and clear audit trails to minimize both frequency and severity. Post-incident protocols will help to contain financial losses. Like cyber, insurers and brokers will shape the standards for testing, validation, and operational oversight. By linking AI assurance to premium levels, insurance can become a catalyst for safer, more trustworthy AI adoption, rewarding those who invest in resilience and transparency while discouraging reckless deployment.

This article first appeared on Instech.

The recent VIU by HUB Personal Insurance Marketplace Report and Rate Guide shows how broader economic trends affect premiums. Notably, the findings demonstrate that standard home and auto rate slowdowns are linked to cooling inflation, while rising rates for homeowners in disaster-prone areas stem from the increased frequency and cost of natural disasters.

With that data in mind, let's break down three of the biggest takeaways from our report:

1. Consumer expectations are rising amid rate fluctuations

Consumer expectations for comprehensive, affordable coverage are rising amid continuing rate fluctuations. While rate hikes may be stabilizing, volatility persists. Tariff announcements earlier this year created short-term disruption, but pricing has begun to level out; auto insurance premiums are moderating slightly, with the average increase closer to 10%.

Not surprisingly, on the property side, homeowners in catastrophe-prone areas still face steep increases. Regions at risk of wildfires, hurricanes, hail and convective storms will likely continue to see double-digit rate growth.

Flood policy counts are also rising as both public and private coverage evolves. As risks expand beyond traditional flood zones, pricing is increasingly tied to localized data.

All these factors have contributed to consumers feeling frustrated not only by rising costs but also by the complexity of an evolving insurance landscape. Insurance is no longer just an add-on to a home or car purchase – its cost makes it a major financial decision. This frustration is driving greater demand for clarity on coverage and costs, supported by neutral, expert guidance.

2. Embedded insurance shows customers that businesses care

Because insurance now occupies a larger share of consumer budgets, it's further affecting consumers' ability or willingness to make discretionary purchases. As a result, business leaders are asking: How can we help customers easily secure adequate, cost-effective coverage they feel good about?

The answer: enable customers to access insurance options through a licensed digital brokerage, either at or after point of sale, by embedding that brokerage's platform and expertise into a brand's sales process. At VIU by HUB, we call this brokerage as a service (BraaS). By partnering with a licensed brokerage, businesses can offer their customers multi-carrier solutions, paired with expert guidance, as part of the customer buying journey.

Think about it: when someone buys a car or applies for a mortgage, they're excited about their new car or home - insurance isn't top of mind. They're navigating a life event, and when it comes time to find insurance coverage, they likely need help. Embedding insurance options and expert advice before, during or after that purchase process is a convenient follow-through solution. For example, we recently worked with a top global automaker to integrate a digital insurance brokerage experience into its customer journey, giving customers a fast, trusted way to compare rates and receive licensed guidance across auto, home, motorcycle, renters and more.

The BraaS model places insurance within trusted shopping experiences and supplements it with live advisors, easy-to-understand choices and seamless operations. This doesn't just reduce confusion. It elevates customer satisfaction, increases revenue and brings long-term value to businesses seeking loyalty and repeat transactions.

3. The blend of human insight and AI efficiency benefits everyone

Artificial intelligence is evolving quickly and reshaping nearly every industry. Insurance brokerages are exploring how AI deepens understanding of customer needs to improve service, without sacrificing the critical element of human dialogue. We believe that, in the near-term, AI is most valuable when it enhances rather than replaces the human experience. For example, AI can be used to accelerate service by flagging life changes or analyzing documentation. It can also be used to give consumers more options and accessibility, such as after-hours service. But the role of empathy, judgment and trust cannot yet be replicated by algorithms. The future isn't human vs. AI – it's humans and AI. Our interactions use technology to support and enhance the human connection where it makes sense and in a way that consumers prefer.

What Comes Next

There's good news: price stabilization is beginning to take hold across some insurance sectors. Carriers are re-engaging in disaster-prone markets, even as rising claims costs, weather-related losses and increased repair costs remain real challenges.

The biggest shifts are consumer-driven. The insurance experience of tomorrow will be defined not by carriers' direct efforts, but by businesses meeting consumer expectations for convenience, clarity, compassion and capability. Embedded omnichannel brokerages are a powerful way to better serve those needs, enabling customers to access trusted insurance solutions as part of making the biggest purchases of their lives.

To better understand how insurers are addressing climate-related changes, the National Association of Insurance Commissioners (NAIC) requires insurers to file a Climate Risk Disclosure Survey that provides a discussion of how companies are addressing these risks and opportunities and integrating them into their governance structures, strategies and risk management.

We have reviewed many of these surveys. What is quite clear, albeit not surprising, is that insurers recognize that they face increased risks from chronic and long-term changes in climatic patterns but, importantly, they are well organized to deal with continuing and impending threats.

What is also clear from the disclosure, and perhaps less appreciated, is that insurers also foresee transformative opportunities. Specifically, these opportunities are expected to come from policies with risk control and loss prevention services. These policies help clients mitigate and build resilience to physical and climate risks. They also see underwriting opportunities in evolving and emerging technologies and industries.

Since insurers are investors as well as underwriters, they also see opportunities in allocating capital toward the energy transition and decarbonization efforts that occur in response to climate change.

The risks are well known but can be managed

Insurers are exposed to a wide array of physical and transitional risks stemming from climate change.

The physical risks arise from the increased frequency and severity of extreme weather events. These events, in combination with population growth and various socioeconomic factors, are likely to keep payouts climbing.

In addition to physical risks, insurers are faced with potential governmental responses to climate change that create conflicting policies across jurisdictions. This could lead to restrictions on insurers' ability to manage exposures and mandated coverages. There could also be higher compliance expenses, and costs for managing additional regulatory standards, especially those that require more disclosures.

Together these risks cannot be downplayed because they can have significant financial implications for insurers. However, the risks are well known, and managements have the right strategies and tools to deal with the issues.

But insurers also see underwriting and investment opportunities. They are expected to come (1) from policies with risk control and loss prevention services that help clients mitigate losses and build resilience, and (2) from coverages for emerging technologies and industries. Since insurers are also investors, they see opportunities in allocating capital toward energy transition and decarbonization efforts.

Mitigation and resilience

With higher climate-related losses in the future, mitigation will be critical to maintaining coverage availability and affordability. To do that the insurance industry will evolve from one that simply pays claims to one that has a critical role in helping policyholders reduce losses.

Hence, the key opportunity will come from designing policies with financial incentives to change behavior. This will be done not just for individuals and corporations, but also at the community level. Some resilient home incentives we expect to see expanded in the future are fire-resistant improvements, and elevated buildings in flood zones, among others.

In addition to offering products to reduce losses, insurers also see opportunities in designing coverages to enhance the ability to recover and adapt after events occur. Thus, insurers foresee the continued development of parametric insurance tied to climate indices (e.g., rainfall, temperature, and wind speed) that enables rapid payouts post events.

In addition to underwriting products, insurers see opportunities in offering comprehensive risk control and loss prevention services. These services include evaluations, technical information, consulting solutions, and educational resources to help clients reduce exposure to physical and climate risks.

Coverage of the latest technology

Insurers (as well as many others) expect that in response to climate changes there will be a gradual transition to clean energy technologies, infrastructure, and processes which will require insurance coverage.

These are some examples of what insurers note in their surveys:

• Renewable Energy Insurance: Coverage of onshore and offshore wind farms, solar farms, green hydrogen facilities, battery storage, hydroelectric plants, and energy conversion risks.
• Climate Technology Insurance: Coverage for climate technology developers, EV charging stations, and suppliers of solar panels and photovoltaic inverter solutions.
• Green Building & Construction Insurance: Product tailored for LEED®-certified construction, green upgrades in buildings, and construction projects focused on climate patterns like flood control, waterproofing, and fire safety.
• Electric Vehicle (EV) Insurance: Developing products and services to meet EV owners' needs, including specific EV policies, roadside charging coverage, and discounts for hybrid/electric vehicles.

On the investment side, insurers see climate change as an opportunity to generate attractive, risk-adjusted returns by deploying capital toward the global transition to a low-carbon economy, supporting emerging climate technologies, investing in green and municipal bonds, and enhancing community and infrastructure resilience.

Insurers anticipate that opportunities will continue to arise related to innovative technologies and solutions across a wide range of asset types. Among others, this includes support for climate change infrastructure, clean transportation, green buildings, pollution prevention, and sustainable water and wastewater management.

Conclusion

There is a natural tendency to view climate change negatively for insurers. But the world is adapting, albeit at a sometimes inconsistent pace, and the insurance industry has the opportunity to take a leadership role in this transition.

The bottom line is that insurance is a risk transfer business. Greater risks can lead to higher growth and enhanced returns if properly managed.

The Co-Op's recent cyber attack, costing the organization an estimated £206 million, is an urgent reminder that no amount of cybersecurity spending can guarantee total protection. Despite substantial investment in defensive technologies and training, a single sophisticated social-engineering attack was enough to trigger a major financial loss.

For many UK businesses, particularly SMEs, the Co-Op incident raises a concerning question: if a national giant with advanced defenses can fall victim, what chance do we have? Yet, according to industry data, one in three SMEs still operates without any cyber insurance. The misconception that strong cybersecurity alone is sufficient continues to leave firms dangerously exposed.

The evolving threat landscape

Cyber threats are no longer static, nor are they confined to traditional realms such as ransomware or data theft. We're now seeing the rise of AI-powered phishing campaigns, adaptive malware and real-time dark web trading of stolen credentials. These tools evolve faster than most defensive technologies can keep pace with.

Human-centric manipulation tactics are becoming more sophisticated and more complex to spot thanks to readily available Gen-AI tools. Malicious actors no longer need to breach technical barriers when they can simply influence an employee to grant access. The recent surge in Business Email Compromise (BEC) attacks exemplifies this trend: a persuasive message, an urgent tone and a brief lapse in vigilance can bypass even the most advanced security systems.

As these social engineering techniques evolve, conventional perimeter-based security measures are increasingly inadequate. The most vulnerable component is rarely the software itself, but rather the individual operating it. Stopping the breach is only the first step; protecting the business's financial stability that follows is just as critical.

Lessons from the Co-Op

The Co-Op incident illustrates a harsh truth: resilience can be incomplete without a financial safety net. Despite its advanced infrastructure, the organization lacked comprehensive cyber cover. That absence meant the full cost of response, recovery and reputational damage fell squarely on its balance sheet.

This is not an isolated case. Many businesses, large and small, still view cyber insurance as optional or redundant, something to consider after implementing technical controls. But as the Co-Op's experience highlights, even the best security architecture can't always account for human error, insider threats or sophisticated deception.

Cyber insurance is there, in the event of a breach, to support a business in minimizing the impact of associated losses. Best-in-class policies can cover everything from forensic investigations and legal costs to lost revenue, data restoration and communications support. In an environment where the average UK cyber incident now costs £10,830 for SMEs and well into the millions for larger firms, that safety net can be the difference between recovery and collapse.

The Trojan horse problem

Think of cybersecurity as the walls of a fortress, essential, strong and well-maintained. But history shows that many fortresses have fallen not because the walls were weak, but because someone unknowingly let the enemy in. A single misplaced click, a compromised supplier or an outdated plug-in can act as a modern-day Trojan horse.

Even companies with specialized IT departments, advanced monitoring systems and extensive backup plans are susceptible. The combination of opportunism, psychology and technology poses a threat that goes beyond simple external factors. Furthermore, when an incident happens, continuity, trust and cash flow are all at risk in addition to data.

Cyber insurance: a partnership, not a replacement

The narrative shouldn't be cybersecurity or insurance, but more like cybersecurity and insurance. The two are partners in resilience, not rivals. A well-designed cyber policy complements technical defenses by absorbing the financial shock that follows a breach, while also providing incentives for strong security controls through lower premiums and enhanced underwriting confidence.

Leading providers work closely with clients to provide risk management support, staff training and incident response planning. There are also policies that champion a collaborative model to ensure businesses aren't just insured, but genuinely prepared.

The road ahead for SMEs

For UK SMEs, the takeaway is clear. Cyber resilience is not just a technical issue; it's a financial and strategic one. With margins already stretched by inflation and economic uncertainty, few small businesses could absorb even a fraction of the Co-Op's losses.

Yet, too many still dismiss cyber insurance as an unnecessary expense. In reality, it can be the final line of defence, the parachute that ensures survival if prevention fails. As threat actors continue to evolve faster than any software patch can keep up, combining robust cybersecurity with comprehensive insurance is no longer optional. It's essential.

When it comes to cyber risk, perfection doesn't exist, only preparation does – that preparation must include both protection and recovery.

The unseen forces of AI and cloud computing are never out of the news, yet behind the headlines lies a story of growth and innovation as tangible as bricks and mortar. The heavy computing power required by AI workloads, and the growing global demand for AI technologies, have seen a building boom take place around the world as developers scramble to build the facilities required to meet these needs.

According to market research, up to $7 trillion will be spent on data centers by 2030 – a huge sum driven largely by technology companies in the US and China, while Europe lags a few paces behind. The tech industry's big three, Amazon, Microsoft and Google Cloud, accounted for almost two-thirds of global cloud revenue in Q2 2025. Combined with Chinese companies such as Alibaba and Tencent, their capital expenditure budgets for 2025 reach hundreds of billions of US dollars, much of it geared toward the industrial scale infrastructure and dependable energy sources that high-performance AI and cloud computing now demands.

The latest Allianz Commercial report, The Data Center Construction Boom, explores the extent of this global buildout and questions whether the building bonanza can last. Despite the expansion, several factors could limit growth, including the surging costs of construction. These have escalated dramatically from $200-$300 million to projects exceeding $20 billion. According to Allianz Commercial construction experts, average-sized facilities now cost between $500 million and $2 billion. Along with higher construction prices, the complex nature of data center construction and operation requires specialized insurance coverage for risks such as power supply concerns, faulty workmanship, fire or natural catastrophes.

Data centers are fueling the construction industry

A global buildout is underway to construct the infrastructure needed to support the digital economy. The US will be the largest market for data centers, covering about two-thirds of the total global data center power demand, with 81 gigawatts (GW) by 2028, while China's data center market is building out equally aggressively. Greater Beijing alone now accounts for roughly 10% of global hyperscale capacity. Europe is trailing behind the two superpowers but is experiencing a 43% annual increase in pipeline activity, with London and Dublin as the largest markets (each with over 1GW capacity), followed by Amsterdam, Frankfurt, Paris, and Milan.

The bigger data centers have a huge footprint. The scale of a $20 billion+ facility can involve tens of thousands of workers on site at peak times, with significant equipment and building supplies moving in and out. Timings can be tight. This requires expert coordination, as any missteps or faulty workmanship can lead to potential losses or costly delays.

Data centers' unique risk profile

Building a data center is a complex, multi-disciplinary undertaking, which presents a multitude of risks. One of the main issues is the soaring power demand that threatens to outpace grid capacity and infrastructure. The electricity demand from data centers worldwide is set to more than double by 2030, to around 945Twh. This is slightly more than the consumption of the whole of Japan today, with its population of 124 million.

To avoid power issues, which are the main source of significant outages, with 45%, data center operators are increasingly seeking to reduce their reliance on the grid by generating their own power onsite, including renewables, gas, and even potentially small nuclear reactors.

Fire, heat, and water are also significant risks for data centers, potentially leading to severe property damage or business interruption losses. Lithium-ion batteries are increasingly being used as backup electricity storage. The fire risk associated with these batteries is well documented, particularly in relation to electric vehicles and charging infrastructure.

Large data centers have increased cooling requirements that could drive up water and electricity demand. This may alter the risk profile of data centers and could contribute to an increase in construction and insurance costs. Clients need to work with an experienced team of underwriters who know the business and can support the project from beginning to end, including multi-year coverage and policy extensions as needed.

To read the full Allianz report, please visit https://commercial.allianz.com/news-and-insights/reports/data-center-construction-risks.html

Over the past two decades, building technology for complex, regulated industries, I've seen how legacy systems can persist long after the problems they were designed to solve have evolved. Workers' compensation is one of those systems. Built for a different era, it now faces a modern workforce shaped by hybrid schedules, shifting risk profiles, AI integration, and rising expectations from both employers and workers.

Workers' comp needs today go beyond incremental upgrades, calling for a strategic shift. The tools we rely on must reflect how people actually work: connected, digital, intelligent, and adaptable. That means platforms built on cloud-native infrastructure, automation that reduces rework and lag, intelligence that drives decisions, and real integration across every part of the claims lifecycle.

A System Under Pressure

While overall claim frequency has continued its long-term decline, dropping 5% in 2024 alone, according to the National Council on Compensation Insurance (NCCI), this doesn't necessarily signal improved efficiency. In fact, claim severity rose 6% in the same year, pointing to greater complexity and resource demands per claim. The trends vary widely by industry. NCCI data shows remote office workers continue to see lower claim rates, while sectors like private education have experienced increases, particularly related to workplace violence. Restaurant-related claims declined in both 2022 and 2023, whereas other hospitality segments remained unchanged.

This uneven landscape challenges systems that were never designed for such variability. The result is delayed resolutions and missed opportunities to prevent disputes or speed recovery.

The Case for Cloud-Native

Legacy infrastructure, while dependable in its time, now limits the adaptability required to meet today's evolving demands. In contrast, cloud-native platforms enable continuous improvement, universal access, and scalable performance. They reduce the friction of system upgrades, allow faster deployment of new capabilities, and offer enhanced data security and disaster recovery, essential for industries handling sensitive health and personal information.

When cloud-native architecture is combined with AI-driven automation, the impact becomes transformative. Claims processing can be streamlined end-to-end, accelerating settlements and improving accuracy. Regardless of role or organization, stakeholders across the claims lifecycle gain real-time access to shared data, enabling faster, more confident decision-making. Automated, auditable workflows ensure that key actions are completed without delay, minimizing errors and improving the overall experience for injured workers. Only a modern cloud foundation makes this secure, coordinated collaboration possible.

How Automation Unlocks Productivity

According to McKinsey, insurers that adopt end-to-end claims automation can reduce operational costs by up to 30% while improving claim settlement times and customer satisfaction.

By automating routine administrative tasks, professionals can redirect their time and energy toward higher-value work: applying judgment, building trust with injured workers, and resolving complex issues more effectively. Industry adoption is already underway. A recent survey by Guidewire and Celent found that over 60% of P&C insurers invest in claims automation technologies, prioritizing automated triage, intelligent document processing, and workflow orchestration. These tools reduce cycle times, eliminate rework, and enable consistent handling of high-volume claims. Some insurers have reported measurable gains: 20–30% faster settlements, 15% fewer processing errors, and stronger claimant satisfaction scores.

Making Data Work Smarter

Workers' comp generates immense data, but far too little informs real-time decisions. The next evolution involves applying intelligence to that data in practical ways. Predictive alerts can flag claims that are likely to escalate. Benchmarking can help leaders allocate resources more effectively. Pattern recognition can pinpoint the root causes of common delays.

More innovative data use means fewer surprises, more consistent outcomes, and better support for injured workers and the teams managing their claims. New actuarial models like Mixture-of-Experts (MoE), highlighted by the Casualty Actuarial Society, show how blending different predictive approaches can improve forecasting for claim frequency and severity. In parallel, academic research points to the value of connecting claims data with nontraditional sources to generate more accurate and timely loss estimates.

Why Integration Matters

Workers' comp involves a broad network of employers, TPAs, healthcare providers, attorneys, and vendors. Without integration, information gets trapped in silos, increasing errors and reducing transparency. Integrated systems, by contrast, allow for synchronized data, faster updates, and a more consistent experience for every stakeholder.

This also enables better accountability. When the entire claims process is connected, it's easier to track what's working, identify inefficiencies, and continually improve.

Looking Ahead

The way forward is clear. We need to build a workers' comp system that reflects the realities of today's workforce and the demands of tomorrow's economy. That means designing platforms supporting clarity, speed, coordination, and trust across every step of the claims process.

The system's pressures aren't easing, and expectations are rising. However, the opportunity ahead of us is significant if we align our strategy with the tools and insights already available.

We're not waiting for the future of workers' comp, we're building it now. The time for cautious upgrades is over. Let's deliver systems worthy of the people they serve.

Following last week's elections in the U.S., it seems that for every two pundits you find three opinions about what the elections tell us about the prospects for the midterms coming up in a year. But seemingly everyone has settled on a theme that politicians must and will hit hard: 

Affordability. 

That word is everywhere, and not just among Democrats, who used it to win elections for governor in New Jersey and Virginia, for mayor of New York City and for a host of down-ballot offices, including in areas thought of as Republican strongholds. President Trump has vowed to wrest ownership of the affordability theme from Democrats, and a high-profile acolyte of his running for governor of New York says affordability will be her watchword, too — as does the current governor, who will represent the Democrats. Everybody loves affordability.

No matter who wins the arm wrestling match, you can be sure that billions of dollars in ads will run in the next 12 months hammering home the need to make goods and services more affordable in the U.S.

That represents an opportunity for the insurance industry. 

If you want to understand more about how hard politicians will push on affordability for the next 12 months, here are two good articles: 

• The Atlantic explores how Democrats turned affordability into victories across the board, adapting the core theme for a wide variety of offices in a host of different political environments. The article warns that the winners now have to live up to their ambitious promises, which will likely be hard to do. "Politics isn’t just about the words you put on your bumper stickers," the article cautions. "It’s about what you do if the bumper stickers work." But the writer predicts that affordability will show up in just about every imaginable form in Democratic campaigns across the country.
• The New York Times documents how Trump, and by extension the whole Republican party, has repeatedly claimed that everything is already affordable. Even though economic cheerleading by the Biden administration and the Harris campaign failed to convince voters in the 2024 presidential election, Trump has strained to create an alternate reality where grocery prices are down (they're up 3.1% in the past year), where inflation has disappeared (it's 3% in the past 12 months and has been rising, albeit slowly), where gasoline prices are plunging toward $2/gallon (they're down a penny since Trump took office, at $3.08/gallon, and show no signs of plunging), etc. He will surely continue making up numbers — he always has  — but since last week's elections, the Times says, Trump has declared that "affordability" is "a new word" and has vowed to claim it for his own. Buckle up.

I think insurers can ride this wave.

That may seem like a stretch at a time when rising car prices and disrupted supply chains have combined with some dangerous driving habits to send auto claims soaring and when natural disasters have exacerbated similar inflationary issues for homeowners insurance. There are obvious limits to what the industry can do to make insurance more affordable. Rates have to be sustainable. 

But policyholders are already sensitized to rising insurance premiums, and the drumbeat will continue. Just in the past week, the National Association of Realtors reported that the average age for a first-time home buyer in the U.S. has soared to 40, from 33 in 2020 and 28 in 1991, and insurance premiums are part of the problem. The flood of advertising and attention by politicians in the next year will, if I'm remotely correct, heighten the concerns. So the industry can either ride the wave or be swamped by it. I vote for riding, both by creating messaging about helping policy holders manage costs — and then by doing it. 

For carriers, maybe the best approach is to emphasize the Predict & Prevent model as much as possible. Homeowners insurers can subsidize sensors that detect the potential for fires and water leaks, or even offer them for free if that makes economic sense — as it increasingly does. Carriers can alert people to anything about their properties that increases exposure to wildfires, floods or storms, while offering discounts if the homeowner takes appropriate action. Even outside the insurance equation, carriers could offer advice on preventive maintenance or anything else that could help people manage down costs. 

Auto insurers could emphasize programs to help people drive safely (and lower premiums), to prevent theft, to get the car under cover as a hailstorm approaches, and so on. Anything to show customers that we're on their side as much as we can be.

Claims departments and third-party administrators could work even harder than they already are to expedite payments — and let the world know about their efforts — because people are feeling financial pressure. If there's a way to advance even a partial payment, that would help some people a lot.

Agents and brokers are where the rubber meets the road. They can pay even more attention to financial anxieties among clients and help them maintain the coverage they need while minimizing increases in premiums. 

Those are just some very rough ideas. I'm sure you have better, more sophisticated ones than I do.

My point here is just to note that affordability is going to be a hot button for at least the next year, if not longer. We should get in front of the issue.

Cheers,

Paul

P.S. There's an old joke about a young politician asking an aging veteran what the secret of his long-term political success has been. 

"Sincerity," the old politician replies. 

"Once you learn to fake that, you've got it made."

To be clear, I'm not suggesting faking anything. A) It's the wrong thing to do. B) The industry couldn't get away with it. 

I'm saying insurers should do whatever they can to address the affordability problem — then take all due credit.

Paul Hill currently serves as President at Job Search Intelligence, where he leads development of salary-data tools and analytics for job-seekers, employers and higher-education institutions. With over 20 years of experience, Paul has built a strong foundation in product development, marketing and data services, bringing a strategic focus to market expansion and customer engagement across the employment-intelligence sector.

Marty Ellingsworth is President of Salt Creek Analytics and Strategic Advisor to Pinpoint Predictive, bringing more than two decades of experience in insurance analytics, data science, and risk selection. He has held senior roles at J.D. Power, Verisk Analytics, Allianz, and USAA, where he focused on transforming data into strategic insights that improve underwriting accuracy, claims performance, and customer experience across the insurance value chain.

Paul Carroll

The creditworthiness of a person factors into the risk rating that insurers develop and use to price policies, and you have an intriguing new stream of data that I think could make those risk assessments more precise. We’ll get into how that works, but let’s start with how you developed your model and gathered the data that feeds into it.

Paul Hill

We started in 2008 by collecting compensation data and wholesaling it to compensation consultancies. We sold data on starting salaries to the academic community so they could help students build career plans.

By 2010, we became known for having reliable data on students’ employment and income after graduation. We contracted to do actuarial work for student lenders because we had just about the only reliable data on young people's capacity to repay their student loans. 

We expanded to credit card debt and auto loans and assessed the wealth-building capacity for young people, looking not just at their earning capacity but at household formation and their investment plans. 

This was all rooted in their academic competency.

We saw that trajectories vary significantly. Nurses, for example, represent extremely reliable individuals from a risk perspective. Women who study computer science are very reliable, while men in the same field might hop from startup to startup, making them less predictable. On another spectrum, young people in trades like electricians, plumbers, or HVAC technicians are eminently employable with no student debt, building wealth in their early twenties, with phenomenal financial outcomes.

Complementing the data on earnings potential, we built a model called "How America Spends." We analyze the 135 million households in the U.S., breaking them into income classes starting with households earning less than $15,000 annually, up to those earning over $200,000.

We've collected finely grained data on their annual expenditures—rent/mortgage, food, energy costs, and so on. Insurance, interestingly, we always categorized as non-discretionary (you need it to drive a car), but with years of inflation, more people than ever are treating it as a discretionary expense.

We total all expenditures by different household categories and build models relating to debt burdens to understand their capacity to meet expenses. Half of households are spending 100% or more of their income every month, and understanding how financially constrained households are allows us to estimate what they'll cut back on. We can identify specific trends, like middle-income households reducing certain food categories or cutting back on dining out in favor of eating at home - a trend we observed early in the inflationary years of 2021-2023.

For insurers, we've developed what we call our "default trajectory model." We break out these same households into income categories, age classes, and geographic regions, and our model shows how members of each category respond when facing financial pressures - their capacity to service typical obligations like purchasing auto insurance, maintaining their car, and paying credit cards. 

This model enables us to view how consumers evolve over time regarding credit card debt servicing, car maintenance, maintaining certain levels of insurance coverage, or dropping coverage completely. It's tailored for auto insurance companies to understand their customer base's financial stability and predict potential coverage changes.

Paul Carroll

I could see benefits both in terms of pricing risk and in terms of deciding who to target with marketing efforts.

Marty Ellingsworth

The customer insight and prediction frameworks of customer lifetime value are critical in both marketing and insurance. When you identify a customer who's on a growth trajectory that's going to be stable, who’s going to pay their bills and remain loyal long-term, you've found better risks. They're going to take care of their house, car, and finances. These customers aren't immune to accidents, but they're not reckless either.

The ability to understand how your book is performing in small business and households is valuable—knowing which households will be the most conscientious, which will be the most stable from income, payment, maintenance, and insurance perspectives, and which ones won't.

It's not that you can predict what one person will do, but a million people tend to follow similar patterns. Looking at risks in tranches—by state, class code, or household—reveals important trends. Some households are being disintermediated, some have higher layoff probability, and others show default trajectories. The trajectory analysis is particularly clever because it helps understand how leading indicators connect to subsequent events, revealing the path that problems typically take.

From that same ingredient technology—analyzing people living in houses, driving cars, working in various occupations and businesses—you can draw important conclusions. People with high equity in their home don't stress as much, so even when something bad happens, they behave differently than someone living on the financial edge. You also see patterns in moral hazard, both occupationally and in insurance contexts, where rationalization, opportunity, and motivation influence poor decision-making around fraud or crime.

I'm not suggesting that losing your job makes you a criminal, but it does create motivation to secure money. Sometimes that means finding another job, sometimes changing insurance coverage, and sometimes dropping coverage altogether just to survive. This type of quantifiable analysis is extremely valuable for describing what's likely to happen to your book of business and your overall business performance.

Everyone knows the trend will be downward for someone experiencing income compression, but they don't know by how much, and they can't segment the risk by different types of households. That's precisely what this analysis accomplishes.

Paul Carroll

I imagine this data can shed some light on the prospects for businesses, too.

Marty Ellingsworth

If consumers are losing, the businesses they frequent will be losing, too. You can expect that the total addressable market of businesses that serve a failing consumer will fail, too. You'll just find out that the businesses you thought you were writing just don't renew, and you don't see them. You won't know why you lost them as customers, but they might have gone out of business.

Paul Hill

Exactly. Many of our clients are caught kind of by surprise by outlier behavior from their customers. A lot of insurance companies, for example, have had a surprisingly high drop rate.

For us, these issues aren't so surprising because we're looking at a 20-year pattern by occupation of how people behave over time as they get out of college and into the employment market.

Understanding their capacity to remain employed is critical. When we look at income services and all expenses, we find that 60% of material defaults are a consequence of job loss. 

So understanding employment stability provides great insights as to a person's ability to service debts or just their standard obligations such as buying insurance.

Paul Carroll

Thanks, Paul and Marty. This was super interesting.

A subcontractor's certificate of insurance (COI) looks perfect on paper. Coverage boxes checked. Dates aligned. Everything appears in order—until an exclusion buried deep in the policy triggers a gap that sends liability straight back to the hiring company.

That moment—when "covered" becomes "exposed"—is when most organizations discover underinsurance. And it's happening more often than ever.

Underinsurance isn't just about a business lacking its own protection. It's increasingly about underinsured partners, vendors and suppliers, whose coverage gaps ripple outward, creating systemic exposure for the organizations that rely on them. Rising premiums, budget pressure, and dense policy language only accelerate this quiet threat.

Why Underinsurance Keeps Growing

Economic strain remains a top driver. Smaller businesses often choose insurance based on affordability rather than adequacy, selecting policies that look good on the cost column but leave dangerous gaps in protection. Most policyholders aren't insurance experts, and the buying process rarely offers the clarity needed to make informed choices.

Studies show 75% of small businesses are underinsured, and over 70% don't fully understand what their policies cover. That confusion leaves them—and by extension, their partners—exposed when contracts or project scopes evolve.

For larger organizations, this creates a blind spot. Contract clauses may spell out required limits and endorsements, but without consistent verification, those requirements often amount to little more than words on paper.

The Burden of Manual Compliance

Traditionally, insurance verification has been a marathon of manual work. COIs arrive via email. Someone checks dates, endorsements and entity names. Renewals are tracked in spreadsheets. Exceptions pile up.

Each step is a chance for error—a mismatched name, a missing endorsement, an expired limit quietly lingering in a file until a claim brings it to light.

Outsourced or "concierge" models were once the best path to confidence in compliance. They brought expertise, structure and relief at a time when most companies lacked the internal bandwidth or knowledge to manage insurance risk effectively. Over time, limitations for those models became clear. While they eased some of the pressure, they often shifted the burden rather than eliminating it. Companies still managed follow-ups, vendors still juggled multiple points of contact and compliance teams remained buried in inbox chaos, performing invisible work that rarely got easier over time.

Manual processes move at human speed. Risk, unfortunately, doesn't.

The AI Advantage

For decades, COI tracking lived in the administrative shadows—slow, repetitive, and error-prone. Even outsourced solutions often relied on teams of people manually reviewing documents.

AI changes that.

Modern AI-powered compliance platforms act as the expert partner compliance teams have always needed. Instead of flagging a problem for human review, AI can instantly read, interpret and validate insurance documents—line by line—against contract requirements.

It determines whether a COI is compliant, identifies missing endorsements and delivers clear, real-time next-step guidance to subcontractors. That means projects don't stall waiting on paperwork. Compliance teams don't spend hours in reactive review mode. And risk management finally keeps pace with the business.

Here's how AI enhances every step of the process:

• Real-time validation of COIs and endorsements—no bottlenecks, no backlog
• Automatic interpretation of policy language, turning "legalese" into plain language for easy action
• Instant feedback to subcontractors so gaps close faster
• Predictive insights that identify trends before they trigger losses

By replacing manual review with AI precision, organizations gain consistency, accuracy and speed while freeing compliance pros to focus on higher-value strategy and vendor relationships.

The result isn't just efficiency. It's resilience.

Practical Benefits for Organizations

Companies using AI-enabled compliance are already seeing measurable results:

• Audit-ready, always: Compliance data stays up to date year-round, not just during renewal season
• Clarity across teams: Everyone—from procurement to project managers—understands where coverage stands
• Cost and time savings: AI delivers expert-level oversight without expanding staff or outsourcing
• Stronger risk posture: Continuous monitoring identifies and resolves issues before they escalate

These are not incremental improvements—they're structural shifts in how organizations protect themselves.

Lessons From the Field

In high-risk industries like construction, underinsurance doesn't show up as a shock—it shows up as a pattern: policies purchased on price rather than protection, COIs reviewed too quickly to catch critical endorsements and renewals missed amid competing priorities

Each of these oversights can create costly liability not only for the subcontractor but also for the hiring organization.

Technology-driven COI tracking isn't a luxury any more. It's a necessary safeguard in an environment where liability risks evolve daily.

A Call to Action: From Reactive to Proactive

Underinsurance isn't a minor oversight. It's a systemic risk that demands proactive attention.

The organizations best equipped to handle it are those that frame compliance as a strategic safeguard, not an administrative task.

AI doesn't replace professional judgment. It empowers it. By surfacing key details, simplifying communication and enforcing consistency, AI gives compliance leaders the guardrails they need to protect every project, every time.

Closing the underinsurance gap isn't just about avoiding claims. It's about building stronger partnerships, protecting revenue and moving forward with fearless confidence.

Auto liability insurance is mandatory almost everywhere in the United States. Although few motorists question the product's necessity, many are anxious about paying for it due to the growing distrust of private insurers. Many policyholders feel uncertain about whether they can rely on their insurance carriers in their hour of need. Countless horror stories reinforce the perception that insurers would drag their feet when it's time to pay or unfairly deny valid claims altogether — and get away with it.

Significant rate hikes erode any remaining goodwill American adults have toward auto insurance companies. In a car-dependent society, being forced to spend more on an inescapable auto-related expense would leave a bad taste in people's mouths. Charging higher premiums is justifiable and even vital in many cases, but is the insurance industry blameless for the public's prevailing sentiment?

An Epidemic of Excess Premiums

Car insurers had windfall profits in 2020. The pandemic compelled different levels of government to impose lockdowns to help curb the spread of COVID-19, which reduced the number of miles driven. Less vehicular traffic meant fewer car crashes and insurance claims, shrinking insurers' risk exposure overnight.

The business shutdowns and stay-at-home orders caused by the coronavirus outbreak economically affected tens of millions of policyholders. In the spirit of solidarity, auto insurance carriers provided relief to customers.

In April, over 80% of insurers announced that they would rebate policyholders more than $6.5 billion over the next two months. State Farm credited customers an average of 25% of premiums from March 20 to May 31. American Family paid customers $50 for every insured vehicle in April. Farmers Insurance gave a discount of 25%, while Progressive refunded about 20% of all premiums paid for April and May.

In addition, many insurers temporarily paused policy cancellation due to nonpayment and waived late fees in the spring of 2020. Most premium relief programs ended in June, but some providers continued to entertain requests for flexible payment plans or similar concessions on a case-by-case basis. They were particularly considerate of financially distressed customers, especially those furloughed or laid off.

The Consumer Federation of America (CFA) welcomed the news. The nonprofit noted that car insurers would've overcharged customers whose premiums were based on driving 1,000 miles monthly and imposed rates considered unreasonable due to the sharp decline in vehicular accidents had they not returned a portion of their revenue to policyholders.

While the premium givebacks were a positive gesture from the industry, the CFA claimed that auto insurers shortchanged policyholders. In August 2021, the pro-consumer group and the Center for Economic Justice (CEJ) analyzed the insurance providers' premiums and claims data. They found that the companies collected $42 billion in windfall profits and returned less than a third of them to payers.

About $30 billion in excess premiums remained in the insurers' coffers. The CFA and CEJ claimed that the insurance carriers used the money to inflate the payouts to their senior executives and stockholders.

Government Inaction — Regulators Becoming Bystanders

The CFA and CEJ had been vocal about the possibility of excess premiums as early as March 2020. The organizations foresaw that the then-current insurance prices would be too high when governments prohibited people from moving freely across jurisdictions.

Granted, the COVID-19 outbreak was a black swan event, and no underwriter could have predicted that American roads would be virtually empty for an extended period. However, the regulators could have stepped in and ordered car insurers to return some of the premiums they raked in during the early months of the pandemic.

The CFA and CEJ wrote letters to state insurance regulators tasked with ensuring rates aren't excessive throughout 2020, urging them to take action and require car insurers to return more of their profits. However, the warnings were brushed aside. The consumer advocates asserted that virtually all regulators did nothing to stop insurance companies from pocketing excessive premiums.

A Stern Counterstatement

The American Property Casualty Insurance Association (APCIA), the national trade association representing auto, home and business insurers, denied the findings of CFA and CEJ and called them misinformation.

In a statement released a day after the CFA and CEJ's report came out, APCIA Vice President David Snyder said everything about the consumer advocates' analysis was incorrect. Snyder said that most of what the CFA and CEJ label as profit was expenses paid to sell and service policies, handle claims, cover regulatory fees and pay taxes. The vice president added that auto insurer profits accounted for just 2% of each premium dollar.

Snyder explained that the number of miles driven quickly resumed to pre-pandemic levels after governments lifted the lockdowns — only this time, motorists' driving habits worsened. Data from the U.S. Department of Transportation's National Highway Traffic Safety Administration (NHTSA) backed this claim.

According to the NHTSA, fatal crashes jumped by 6.8% in 2020. About 45% of the drivers of the passenger vehicles involved were speeding, alcohol-impaired, not wearing a seat belt or any combination of the three risky behaviors.

A Period of Underwriting Losses

The property and casualty segment of the insurance industry was in the red in 2022 and 2023. Rick Gorvett, a Casualty Actuarial Society (CAS) fellow, described the two-year period as part of the insurance business's multiyear cycle. The market was soft, partially driven by competition and the ebb and flow of alternative risk management mechanisms.

Supply-chain disruptions contributed to the underwriting losses. The pandemic exacerbated the existing chronic semiconductor shortage due to lockdown-related production downtime, consequently limiting new vehicle production. Lower car inventories drove up the cost of new and used automobiles, ultimately increasing the value of assets insurers have to cover. Rising repair costs due to modern vehicle designs added more burden to insurance companies.

Gorvett pointed out that the social factors, like increased litigiousness, inflated insurance losses. A 2024 joint study by Triple-I and the CAS found that auto liability losses and defense and cost containment expenses spiked by $76.3 billion to $81.3 billion from 2014 to 2023 due to the involvement of billboard attorneys in claims and considerable tort awards.

A Loss-Loss Scenario

The U.S. Bureau of Labor Statistics said that vehicle insurance prices skyrocketed by 17% in 2023 year-over-year, outpacing those of food away from home, housing and electricity by a mile.

Such inflation happened at a time when car insurance underwriters were seeking sustainable rate adequacy levels. Considering that a 2019 study revealed that Americans already overspend about $37 billion yearly on auto insurance, the 2023 figure meant that car-related expenses made up a larger chunk of household budgets.

Elevated crime rates also fueled the 2023 premium hike. 2022 was the worst year for vehicle theft since 2008. Data from the National Insurance Crime Bureau showed that thieves stole over 1 million automobiles that year, some of which ended up in the hands of vehicle traffickers operating in foreign countries.

Underwriters had to factor crime rates into the equation. Underwriting decisions affected customers living in hot spots, whether or not they had fallen victim to car theft.

So consumers and insurers both lost.

The Post-Pandemic Premium Surge

Since the pandemic, 2024 has been the most profitable year for personal auto insurance carriers. The industry segment recorded a net combined ratio of 95.3 and posted a 13% increase in net written premiums year-over-year.

Pundits credited the improvement in underwriting performance to pricing realignment. Efforts to align auto insurance prices with soaring risk levels and implement more effective control loss measures paid off.

Low claim frequency was another 2024 highlight. While the number filed remained below pre-pandemic levels, claim severity was high. Shifting attitudes toward car insurance can explain these two phenomena.

The Jerry 2025 State of the American Driver Report revealed that more consumers are motivated to shop around after seeing rates balloon by over 50% over the past three years. In 2024, 55% of motorists sought better deals in the past 12 months, and some changed vendors to obtain more affordable rates.

Moreover, many drivers bought less coverage and agreed to higher deductibles to lower their premiums. These decisions suggest that more policyholders are keener on filing claims to cover larger losses. Prudent consumers know their claim histories can drive up their future rates, so they only want to call their insurer when it counts the most.

Increasing Auto Insurance Rates Legally

The compulsory nature of car insurance policies makes premium hikes unpopular at best and controversial at worst. Policyholders almost always feel the pinch, regardless of whether auto insurance turns a profit. Judging by how effective the legal system can be when compensating the insured, insurers must navigate the regulatory landscape carefully to raise premiums accordingly without breaking the law.