In recent conversations, Brian Poppe at Mutual of Omaha highlighted AI adoption in four stages – a transition from AI being a fun tool that employees use to a final stage where AI-driven processes are integrated for seamless workflows. And certainly directionally, this appears to be correct – AI use cases have historically been focused on addressing specific inefficiencies and then scaling, which will eventually lead to AI driven processes.

While there is an acknowledgment that AI is a rocket that will move insurance in a truly transformative way, carriers are treating AI as if you are putting a new engine in an old car: it performs better, but it is still driven the same way.

AI-driven processes raise an entirely different question – do legacy operating models make sense for insurance carriers in the age of AI? And the answer is – no, they do not.

When we think of an operating model, definitionally we should think of the way in which people, processes, technology, and capabilities are arranged within an organization to deliver value to customers. The evolution of AI will be AI assisting with a task to developing a process that is driven by AI. But operating models are derived from the assumption that processes are human-driven.

Consider the underwriting process and how the evolution of technology has changed it. Historically, you needed many underwriters to carefully review applications, assess the risk, and then provide a quote on pricing against that risk. RPA reduced the manual effort and increased productivity, but the process remained the same. Then automation and enhanced underwriting (e.g., algorithmic, usage-based, simplified, etc.) were implemented to provide faster underwriting, but the organization did not necessarily change to reflect these changes. Instead, carriers have viewed this from the lens of capacity and workforce management.

But if you were building a new insurance carrier today, would you structure underwriting in the same way that it is today? Most likely, no. And as AI evolves over time, you would certainly design a different operating model.

In other words, processes designed around AI and technology would require a quite different organization than human-driven processes. The more carriers lean into AI-driven processes, the more the legacy operating model makes little sense.

The Legacy Trap: Why Current Models Are Not Changing

If we accept that AI-integrated processes are directionally where the insurance industry is headed, then the question is why haven't carriers designed new models? There are several reasons why organizations are not evolving:

1. Organizational Resistance: AI-driven processes come with an uncomfortable question – what is the role of a human in this new environment? Most assume that it means that AI is "coming for their role," and to some extent, they may be correct. But that assumption hinges on two beliefs – that all capabilities can be automated and that all automated capabilities no longer require people. Neither of these beliefs is true.

2. Lack of Success With AI: There is an often-cited statistic that 95% of all AI projects do not make it from pilot to tangible, measurable ROI. This suggests that although carriers are investing in AI and understand its capabilities, they are not finding success at scale, delaying transition to AI-driven processes and capabilities. While this suggests that the AI-driven process may not be as close as some believe, it would be incorrect to dismiss it as hype. Executives and insurance leaders only need to be directionally right, and innovation in the space should be balanced with an AI strategy on what to invest in and how to prioritize.

3. Unproven Models: Insurance carriers are conservative – an op model built on new technology is a significant risk and has not aligned with traditional automation strategies. Typically, a process is automated and then resources are reallocated or modified once the investment has generated ROI. But there is evidence of carriers operating in dual environments with new operating models, in what some have called a "two highway" approach – a legacy environment for in-force business coupled with a new environment for new products. A new target operating model does not need to be an enterprise effort initially – it may be useful to design a different model in a specific business unit to run in parallel to assess strengths and weaknesses before eventually scaling it.

Building AI-Native Op Models: A Practical Framework

If carriers accept that integrated AI processes creating new workflows is the future, then part of the planning effort must be an exercise developing a new target operating model. As carriers seek assistance with developing these models, there are five key principles that will lead to the greatest chance of success:

1. Realize Directionality Is More Important Than Timing: Carriers do not need to know exactly when a transformation will occur, they only need to think in terms of where AI is moving directionally. Consider various capabilities in insurance. Operational support of the insurance model is likely headed toward significant automation of processes, while sales and marketing are likely to remain less automated in the future. From an operating model perspective, that likely means that AI driven processes will push workflow in the back office (think of new business submission or policy administration), while in the sales capability, you are more likely making the agent/advisor/broker more efficient (e.g., next best actions, generating marketing material with existing pre-approved templates).

2. Ignore Biases and Existing Requirements: One of the most difficult aspects of designing a new operating model in general is getting stakeholders to leave "the way it has always been done" at the door. Remember that this is a white space exercise and should be framed as such. For example, policy servicing should initially be thought of in terms of desired customer/agent experience, not how that service is delivered. When framed appropriately, carriers can focus on what they want to achieve and then assess how they would achieve it.

3. Understand the Hard Lines: For some carriers, there are hard rules that they will not consider. For example, risk appetite in underwriting may make some AI-driven processes impossible, or there may be a decision to create a large case workflow that is human driven to provide white-glove treatment for a particular agent class. Understanding enterprise non-negotiables upfront eliminates downstream decision-making on the op model.

4. Embrace Uncertainty: Carriers must understand they are blazing a new path forward. There is no cookie-cutter approach to a new operating model. While there are proven approaches, the result is that you may no longer have a clear benchmark. AI is introducing uncertainty and the only thing that we know is that it will transform the way that insurance carriers operate. The introduction of AI-driven processes will inevitably create a feedback workflow connecting actuarial product design, underwriting, and claims to create real-time adjustments to initial assumptions. The long-term consequences are unknown, but carriers still have to develop these capabilities to compete in the market.

5. Iterate, Iterate, Iterate: While there is directional design, understand that operating models evolve as new data is presented. While there are assumptions that sales (particularly personal lines) will continue to be driven through agents and brokers, significant change in customer dynamics or technology could change these assumptions. Additionally, end-state operating models make assumptions on where technology will be, not where technology is today. That may mean an agile approach to op model development.

The process of developing these operating models will not be instant. But carriers must begin the process of reassessing how they are organized to meet client needs in the age of AI. Digitally native carriers like MGT Insurance (organization built around AI stack to support small businesses) and Ethos (organization built around underwriting that can be done in five minutes) are already further along in this journey than legacy insurers, and the consequences may mean bloated organizations, reduced profitability, and an inability to compete in the marketplace, particularly in price sensitive markets. Embracing AI while ignoring op model transformation is only delaying the inevitable. As AI evolves, what assumptions in your op model might need rethinking?

Growth no longer arrives quietly. It comes with evolving rules and regulations, higher expectations from consumers for seamless service, and less room for operational error. Expansion puts every assumption about how an agency operates under a spotlight.

The agencies that succeed are not growing faster by accident. They are building from the start with an eye toward what it will take to operate as a next-generation agency.

What that looks like in practice becomes clear when you examine how a handful of fast-growing agencies have approached scale over the past year. After years of working alongside agencies as they grow and change, those patterns are hard to miss.

When Ambition Forces the Issue

Look across agencies at different stages of growth, and a pattern emerges. Ambition is rarely the constraint. Execution is. The divide came into focus when we worked with a newly formed agency that entered the market with clear and aggressive growth objectives.

The founders were not new to insurance, but they were clear-eyed about the risks. Rapid expansion without proper structure would create compliance risk, service inconsistency, and operational drag. Rather than treating those challenges as problems to solve later, they treated them as foundational design requirements from day one.

Designing for Scale Before It Is Required

Instead of layering tools and processes reactively, the agency focused on building repeatable frameworks. Compliance expectations were standardized and ingrained into processes and systems. Service models were defined. Training and onboarding were designed to work across locations and teams.

This approach created clarity early. New offices could launch efficiently, without reinventing how the agency operated. Agents could onboard quickly, without sacrificing quality or oversight. Leadership retained visibility as the organization expanded, and could quickly course correct where needed.

The company is now well positioned for continued expansion without the loss of control that typically accompanies rapid growth. The takeaway is not that speed matters most. It is that discipline and sequencing matters. Infrastructure came first. Scale followed.

Why Growth Exposes Weak Operating Models

Many agencies discover their operational limitations only after growth accelerates. Processes that worked at a small scale begin to break. Informal knowledge becomes a bottleneck. Compliance shifts from manageable to overwhelming.

In response, agencies often add more tools. A system for enrollment. Another for compliance. Another for reporting. Each addition solves a narrow problem but increases fragmentation. Over time, leaders lose a clear view of what is happening across the business. Agents spend more time navigating systems than serving clients. These issues create motion without momentum. Focus on the customer inadvertently wanes. Growth begins to slow, and further scale becomes next to impossible.

The Difference Between Scaling and Expanding

There is a meaningful distinction between expanding and scaling. Expansion adds volume. Scaling adds capacity.

Agencies that scale successfully build operating models that absorb growth without degrading performance. Compliance and quality remain consistent. Service delivery is predictable. Visibility improves rather than erodes as volume increases. This requires standardization without rigidity. Processes must be repeatable, but flexible enough to adapt to different markets and consumer needs. Growth becomes something the organization plans for and manages deliberately, rather than reacting to as problems arise.

Rethinking Revenue and Retention

Growth also forces agencies to confront how they think about revenue.

In the case of another agency we recently worked with, which was entering a growth phase, leadership recognized that focusing on short-term results was creating an unstable foundation. Leadership began to prioritize lifetime customer value and persistence as core performance metrics.

Product strategy was aligned with long-term outcomes rather than immediate payouts. Agents were better educated on how coverage decisions affected customer satisfaction over time. Data was used to reinforce better decision-making at the point of sale, with an intense focus on customer satisfaction as key to an effective lifetime value model. The result was a healthier book of business and more predictable growth. Revenue was no longer completely reliant on obtaining new customers. It was supported by durability and lifetime-value-based business objectives.

What This Means for Agents

When workflows are clear and systems are coordinated, agents spend less time navigating administrative tasks and more time working with clients. Expectations are consistent across the organization, support is easier to access, and day-to-day work feels more predictable.

That stability matters. Growth no longer feels chaotic or dependent on workarounds. Instead, agents operate in environments where processes support them, allowing them to focus on building relationships and growing their business with confidence.

Where Agencies Pull Ahead

Growth itself is not a differentiator. In every thriving business, growth is expected. What separates agencies is whether they can scale without losing control, consistency, or trust. The real challenge is not adding volume but sustaining clarity as complexity increases.

The agencies that succeed will not be defined by how quickly they expand but by how intentionally they build for the future. Compliant growth becomes a foundation rather than a constraint, and processes are designed to repeat and scale instead of relying on individual heroics. Growth is not a moment to chase. It is a test of whether an agency was built to last.

Once the carrier's most trusted ally, mainframe systems/on-prem applications have now become outdated. Their prowess compared with modern technologies such as cloud computing and artificial intelligence (AI) appears bleak at best. They're expensive to maintain. According to a BCG Analysis, global IT spending in the insurance industry was about $210 billion in 2023 and is expected to grow 9% annually through 2027. Plus, evolving customer digital appetite, the competitive landscape, and regulatory complexity put perpetual pressure on traditional insurance businesses to make that ultimate call. Application modernization!

That said, complete legacy rewrites remain too risky and costly. These core systems power mission-critical functions, including underwriting, policy issuance, claims approval, renewals, and compliance. Complete rewrites can lead to the loss of critical business logic, reintroduction of old bugs, and increased security vulnerabilities. Most importantly, rewrites can disrupt operations, affecting existing policyholders' trust, revenue, and regulatory standing.

That's why insurance CIOs today appear to be at a crossroads. How can we accelerate modernization without breaking what works and risking downtime? The key here is moving toward a modern core environment with digital capabilities. In this article, we discuss a strategic, outcome-driven transformation approach that helps CIOs introduce modernization into core insurance processes while increasing customer satisfaction and productivity while cutting costs.

What's Happening in Insurance Businesses Today: The Need to Modernize

For years, insurers have persisted with traditional processes for day-to-day operations that have been heavily reliant on paper-based interactions, on-prem legacy systems, and CRM databases. This means that agents spend more time on administrative tasks and less on risk assessment, actuarial analysis, processing claim submissions or bringing innovation. However, over the last few years, technology has evolved at breakneck speed. Businesses are leveling up innovation by bringing in AI across various functions.

Consumer psychology is also being reshaped with each passing day as AI penetrates different spheres of life. People are using AI/LLM tools to generate content, build apps, get new business ideas, seek therapy, create financial plans, and whatnot. The human brain is being rewired to get personalized information in seconds. And as people get increasingly used to instant gratification, the insurance industry, with its limited customer touchpoints, will find it harder to keep them engaged.

Especially, with legacy systems that have become more prone to downtime as the seasoned professionals who have maintained these are aging or already retired. Even regulations have become an issue, necessitating the need for a robust, modern data security infrastructure.

Lastly, it's also about keeping up with the world. To stay competitive, insurers must adopt cloud and new technologies such as generative AI. And for that, CIOs don't need to face operational disruption, cost overruns, and service degradation. Phased modernization approach can prove to be effective.

Initiating Modernization: How Can CIOs Move Forward

Deciding whether to build custom solutions in-house, upgrade existing systems with modern wrappers, or purchase a ready-made platform is a complex decision. Carriers are contemplating this. In the United States, roughly half of the leading P&C carriers opt to buy and configure systems, while the other half decide to build. Each approach has its merits.

1. Transforming Existing Systems Incrementally

CIOs seeking to revitalize their legacy software systems can choose between less invasive approaches, such as code refactoring and replatforming, and more complex transformation strategies, such as rearchitecting. For less complex applications, former approaches work best. Existing core insurance apps can be moved to modern cloud infrastructure with minor adjustments, improving performance while reducing operational costs. That said, this transition can take months using traditional software development methods. In continuously evolving markets, the cost of waiting is more than what businesses can digest. This often appears as a daunting mountain, making IT leaders abandon the idea altogether. This is where a step-by-step approach, incremental modernization, becomes more effective.

It balances business continuity with technical evolution. It's not about updating one app/core function at a time. Incremental modernization is about identifying modular modernization units: workflows, sub-domains that can evolve independently. That means reshaping the system from within.

2. Using AI Agents as Intelligent Wrappers

CIOs can leverage modern wrappers, such as AI agents and RPA, to improve efficiency. These modern technologies can work alongside your primary heritage applications as digital assistants (modern wrappers) without replacing or changing the system itself. These AI agents interact: observe (read), monitor (listen), and act (share real-time insights) while boosting security and reducing technical debt. This can empower human agents to make more effective decisions as they act on real-time insights, helping boost productivity, increase customer satisfaction, and maximize ROI.

3. Simplifying Workflows with a Unified, Scalable Insurance Software

CIOs can also opt for modern insurance software to facilitate quick deployment, shorten modernization timelines, and meet modern business requirements without heavy customization or a rip-and-replace approach. An AI-enabled insurance management software can streamline an insurance policy lifecycle, including onboarding, underwriting, billing, claims, and servicing. But the main challenge is to select software that can actually become a real growth engine. This is where due diligence counts.

Evaluate the vendor's API capabilities and integration experience. Deduplicate and cleanse data (build a solid data foundation). Run pilot tests and increase stakeholder alignment to facilitate adoption. Include the compliance and information security teams in the evaluation process. These measures can help insurers modernize with greater confidence, reduce execution risk, and move with the agility required in today's insurance landscape.

Alternatively, big insurance firms can benefit more from custom-built software, which offers greater customization to the unique needs of the business.

Conclusion

System modernization is a transformative journey, a meaningful opportunity for insurers to reposition themselves as a modern-day enterprise in the eyes of both the workforce and customers. A business that can skillfully orchestrate complex insurance operations while remaining digitally advanced. And for that, CIOs no longer need to choose between protecting mission-critical core systems or embracing end-to-end digital transformation. Incremental modernization, intelligent wrappers, or a well-evaluated COTS offer a strategic, low-risk path forward.

Don't think of the modernization journey as one that dismantles your most trusted allies, mainframe systems/on-prem applications. Instead, it empowers these systems with modern-day capabilities. It helps consolidate up-to-date data in one place, minimize manual and iterative work, increase customer engagement through personalization, and respond faster to regulatory and market changes.

So, the question is: Are you ready to build a resilient technology foundation that supports sustained, long-term growth and builds a competitive advantage? Those who act now with clarity will gain an advantage tomorrow.

Not long ago, managing millions of documents across dozens of databases was a challenge reserved for the largest enterprises. Today, widespread cloud adoption and a more distributed workforce mean that organizations of all sizes are handling vast, growing volumes of data.

The challenge isn't just scale—it's complexity. Much of this data is unstructured, scattered across systems, and increasingly filled with personally identifiable information (PII) and other sensitive details that are difficult to find, manage, and protect.

Furthermore, the cost of data breaches continues to rise. The 2024 IBM Cost of a Data Breach report states that the global average cost in 2023 reached $4.8 million, marking a 10% increase from the previous year and the highest on record. A significant 75% of this increase is attributed to lost business and activities related to responding to the breach.

Allocating resources to prevent data breaches is already challenging—and it becomes even more complex as privacy regulations like GDPR and CCPA continue to expand. These laws require organizations to maintain greater transparency and stronger protections for sensitive data across its entire lifecycle, covering information collected from customers, patients, employees, and even website visitors, regardless of when or how that data was gathered.

The harsh reality is that data loss has become commonplace. Breaches often go unnoticed for months, and meeting compliance standards is becoming progressively more challenging.

How can businesses confront these challenges? The good news is that the process of identifying, classifying, and fixing sensitive data to mitigate risk can be automated. In fact, as the IBM report states, organizations that heavily use security AI and automation saved approximately $1.8 million compared with those that do not.

How AI Is Transforming Data Privacy

In the age of digital transformation, data privacy is now a fundamental concern for businesses globally. Incorporating artificial intelligence (AI) into data privacy strategies is not just a technological step forward; it is essential.

Like many AI applications, the aim is to boost productivity and minimize human error. In cybersecurity, AI supports a SOC environment by aiding threat hunting, incident response, and daily cybersecurity operations. It enhances value by processing data, offering better context to security teams, and automating routine tasks. The main goal is to use AI to boost experts' productivity, particularly when errors carry high costs.

Primary Functions of AI for Data Protection and Privacy

Proactive data management: Unlike traditional systems that react to issues after they occur, AI adopts a proactive approach. Data security solutions that leverage machine learning to scan, categorize, and monitor data continuously in real time help ensure that PII is securely stored and actively protected.

Deep insights and predictive analysis: AI's strength is in extracting meaningful insights from large datasets. It identifies patterns to forecast potential threats and vulnerabilities, enabling businesses to proactively strengthen their security defenses. Tools that use AI to automatically spot anomalies, such as unauthorized access, risky data sharing, improper permissions, and incorrect locations, facilitate quick responses and corrections.

Adaptive learning: This key feature quickly adapts to changing cyber threats. As threats evolve, AI systems analyze new patterns to enhance security and prevent breaches. Advanced machine learning can scan and classify data, learning from observed patterns. AI-powered risk analysis automatically detects PII, understands its usage, and assesses its risk level. As the system encounters new data formats and usage methods, it updates its knowledge to provide more precise risk evaluations.

The Future of Data Privacy With AI

Data privacy is evolving rapidly, and traditional methods are no longer sufficient. With cloud-first deployments, increasingly stringent regulations, and continuous data growth, protection must keep pace with emerging threats. AI introduces a new pace by detecting risks instantly, adjusting to new data patterns, and making decisions that previously took days within seconds. The future focuses on leveraging AI to enhance data security—making it smarter, more precise, and constantly active.

Here's what AI offers:

Real-time protection: This is crucial as businesses shift to live operations. AI's capacity to process and analyze data instantly makes it essential for immediate protection. By autonomously scanning and classifying data, advanced AI ensures businesses can trust that their PII is continuously protected.

Regulatory evolution: As data privacy challenges increase, regulatory frameworks also expand. AI's flexibility enables businesses to easily adapt to these changing rules while remaining compliant with minimal disruption. AI can adjust its monitoring and protection measures to comply with various regulations.

A collaborative approach: In the future, AI and human expertise will work together. AI will manage real-time processing and threat prediction, while security teams focus on developing and executing long-term data protection strategies. Businesses should seek an AI-driven solution that provides the technical tools and integrates smoothly with human-led strategies and decision-making. This allows businesses to leverage the advantages of both: AI's speed and efficiency, combined with the strategic insight of human experts.

Paul Carroll

Based on what I’m seeing at ITL, customer experience has become a truly hot issue in the insurance industry, especially as customers are more willing to shop around. Is that what you’re seeing, too?

Sean Eldridge

Absolutely. With the advent of many of the GenAI and agentic AI solutions that can be customer-facing—such as agentic voice for inbound and outbound calls—we're definitely seeing more interest.

Just to step back, I think "customer" is often too narrowly defined. Companies are just solving for the claimant, or just solving for the policyholder, or just solving for the client in a TPA-type experience. We've always looked at it as an ecosystem—your claimants, your policyholders, your adjusters, your supervisors, your agents, your brokers. How do you not just optimize for one group but look at them more holistically to make sure any CX solutions don't help one group but potentially hurt another.

The industry should think more broadly about how to help both our people behind the scenes, as well as that end user on the other side.

Emily Cameron

If the user has a positive experience, they understand what's going on and they feel good about what's going on, that will lead to fewer phone calls over to the adjuster, less litigation, etc. There are just so many interdependencies throughout the process.

Thinking in terms of an ecosystem and focusing on that claimant experience—even if maybe efficiency is your primary goal for the year—people are starting to understand how everything is interrelated and focus on how we can simplify things to have a more successful process and experience for everyone involved. This has been our mission from the start.

Paul Carroll

Over time, every industry becomes a technology industry. The computer industry used to be simple—if you had a problem, you called IBM. They sold the mainframe, the software, the peripherals—everything. But with personal computers came an ecosystem based on different software pieces, and solving problems became difficult. Vendors pointed fingers at each other. Insurance is even more complicated because there are two levels of customers: the broker or agent as intermediary and the end customer. How do you think about developing a plan that maps out the customer experience in a complicated industry involving lots of pieces of technology?

Sean Eldridge

When we got started six years ago, we made a deliberate decision to start with the problem, not the technology. Before writing a line of code, we spent thousands of hours with carriers, third-party administrators, self-insureds, claims teams and claimants to understand the friction points across the claims process, the service process, the underwriting process—you name it.

For us, that meant building an underlying architecture first that allowed for a wildly extensible level of configurability and interoperability. We see a lot of fantastic technology coming to bear with the advent of GenAI and agentic AI, but those are still point solutions for specific use cases that are just scratching the surface of what's possible. Without great configurability and interoperability, you can’t do a lot.

When you talk about reducing phone tag, clarifying expectations, and getting the right information to the right person in the right system, really special things can happen in our industry. And again, that starts with the problem, not the technology.

Emily Cameron

I like to think of us as a meat-and-potatoes company. It's easy to see a cool feature or solution and say, "I need that," adopt it, and then find it's just not used. So we go on-site, sit next to the adjuster, and see how their current workflow works to make sure that we're actually improving it and not adding work for them. We even go to grocery stores or hardware stores and talk to those employees to get their experience.

To your point about companies becoming tech companies: I think that's really interesting because oftentimes our main competitor is just the traditional phone call and snail mail. There is definitely a technology adoption curve. People resist. They’ve been doing this a long time. They’re used to their flow.

So we started slow, focusing on their first priority: It's difficult to communicate, so let's improve the messaging capabilities. Once people get comfortable with that, they're upset that things are taking so long. So how can we get information back faster? Well, here’s our electronic document solution to exchange forms faster. Or they say, "There are so many systems I have to deal with." So here’s our ETL solution to deeply integrate and pull in information. Once they get their foot in the door, they start seeing the value.

Paul Carroll

What are some specific examples of problems you've identified through this kind of field research?

Emily Cameron

Oftentimes the biggest struggle we've heard directly from claimants is, "I just don't know what's going on." So we get some resources and information over to them right away. We also built an automated intake solution, to make it super easy to report a claim quickly. Then we can immediately send a text or email while we're still processing their claim and figuring out who their adjuster is. This process can take 24 to 48 hours sometimes, and we don’t want claimants to just be kind of twiddling their thumbs and not quite sure what to do.

We've done virtual interviews with adjusters, and they say, "Hey, I'm fine. I'm just doing my job. No issues." Then you sit down next to them and realize they're spending hours a week just looking through files manually, and they didn't even think to bring that up.

Once we build a relationship, they’re more likely to say, "Hey, I'm having this issue. Is this something you guys could take a look at?" But at first, people just don't know what they don't know, especially if they're not used to technology.

Sean Eldridge

I can think of countless examples where we've been in the trenches side by side with claimants, policyholders, and insurance professionals to really understand their challenges. Someone says, "Oh, I have no problem keeping track of all these things." Then you see their desk, and it's got 74 Post-It notes on it. You're thinking, okay, there might be a better way here.

The devil is in the details on how those solutions come to life. There’s been an explosion of AI document processing tools over the last 18 months, but how do you think about interoperability, whether with the document management system, the core systems or whatever? How do you let claims professional configure that document processor but still within guardrails set by the organization? There are all the fine details you don't know until you're in the weeds with those individuals and teams.

Paul Carroll

Configurability has always been a bugbear. I vividly remember the early days of enterprise resource planning (ERP) systems, led by SAP. They were great, but even the biggest customers pretty quickly found that they had to redo accounting, requisitioning and other processes to fit SAP’s way of doing things, when it should have been the other way around.

Emily Cameron

Every company we've worked with likes things a different way. We call them the "special snowflake." So we try to make sure that almost everything is configurable.

It even gets down to the user level. One adjuster told us, "You know, I just don't want text or email. I like the old days when someone called me." So, we built a system that calls that person and uses what sounds like an authentic voice to let them know they have a new claim or an update.

Paul Carroll

I mostly see customers wanting claims to move faster, with regular updates, and you’ve talked about those issues. What are other key touchpoints that define a good customer experience in insurance?

Sean Eldridge

At a high level, it's reduction of uncertainty. Whether you're talking about a claimant versus a policyholder versus even the adjuster or other insurance professional—a client service rep who might be touching that claim or client in some way, shape, or form—everyone just wants to be able to set expectations. When can I expect to hear something? When I have a scratch-my-head type moment, how do I get an answer for it really quickly?

I think the industry is still just scratching the surface of what's possible on reducing uncertainty. Personal lines have probably done better than commercial lines historically, but there's a sea change coming in terms of what that's going to look like.

And I think the research-based approach—problem first, technology next—will get us there. One of our cofounders is a highly published researcher and has a focus on behavioral science. We've got the former lead out of IDEO's behavioral design division who's helped us think about, from a human factors standpoint, how do you ask the right questions at the right time to unlock insights? And I think our approach is part of a rising tide that will lift all boats in the industry.

That's very exciting in terms of improving customer experience for the long run.

Paul Carroll

Here’s hoping. Thanks, Sean and Emily.

 

About Sean Eldridge

Sean G. Eldridge is the Co-founder and CEO of Crosstie, a venture-backed insurance technology company that helps P&C carriers, TPAs, and self-insured organizations modernize claims and service workflows through configurable AI and automation. Outside of Crosstie, Sean led a private equity-backed roll-up in the disaster restoration industry, giving him firsthand exposure to the realities of claims operations, and previously held leadership roles at Johnson & Johnson, Procter & Gamble, and Weight Watchers focused on building and scaling technology-enabled services. He earned a B.S. in Management Information Systems from Rochester Institute of Technology and an MBA from Harvard Business School. He resides in Cambridge, MA with his family.

About Emily Cameron

Emily Cameron is the Head of Product and Customer Success at Crosstie, where she leads the development and adoption of technology that improves claims and service outcomes for P&C carriers and TPAs. By unifying product strategy with customer success, Emily ensures the platform delivers measurable operational efficiency, clearer communication, and better experiences for both insurance professionals and the people they serve. Emily began her career at Epic, one of the world's largest healthcare software companies, where she held escalating technical and customer-facing leadership roles supporting complex, mission-critical implementations for large organizations. She later joined Crosstie to build the product and customer success function as the company scaled its platform across P&C insurance. Emily holds a B.S. in Bioengineering, cum laude, from the University of Washington.

The IoT industry is growing fast, changing cities, factories, healthcare facilities, and industrial sites. To remain competitive, businesses are employing the most advanced technologies that make IoT systems smarter, faster, and more secure. In this article, based on our experience in IoT consulting, I'll explore the key IoT trends that are expected to lead the market in 2026.

1. Sentient AIoT

According to the research published by Markets and Markets, the AI-based IoT (AIoT) market was valued at $25.44 billion in 2025 and is estimated to hit $81 billion by 2030, growing at a CAGR of 26% during the forecast period. Embedding AI into IoT solutions reduces human error through automated daily operations and facilitates real-time data analysis. AI-based IoT systems monitor operational data, environmental conditions, and equipment status, using this information to continuously optimize operations.

AIoT is valuable for various operations and workflows:

• Real-time threat recognition: Detecting abandoned objects or unusual activity, such as crowd formation, movement during off-hours, or any other predefined suspicious behavior, and triggering instant alerts for security teams by using AI-powered cameras.
• Quality control: AI-driven vision systems for production lines can identify defects in manufactured parts, preventing the release of faulty goods.
• Predictive maintenance: Analyzing sensor readings and equipment configuration history, AIoT systems help forecast equipment failures before they occur, reducing downtime and repair costs.

2. Cloud-edge hybrid architecture

Hybrid architectures combining cloud and edge computing continue to increase in popularity for IoT deployments as they address the limitations of cloud-based architectures in terms of bandwidth, scalability, and security.

Processing data close to its source on edge devices significantly reduces latency and enables real-time responsiveness and immediate decision-making. At the same time, cloud servers provide scalable resources for data analytics and storage, data aggregation across multiple devices and locations, and AI model training. In hybrid environments, security can be reinforced through zero-trust architecture, a modern cybersecurity framework, because data is spread across multiple edge and cloud environments, requiring continuous verification for secure communication between endpoints.

Advanced technologies, such as 5G for high-speed connectivity, containerization for flexible deployment, and AI-powered resource management optimization help maximize performance in cloud-edge architectures.

The hybrid architecture is valuable for various applications in smart cities, medical facilities, industrial automation, and autonomous vehicles, providing the base for low-latency IoT ecosystems.

3. Sustainability-driven IoT

The goal of green IoT is to reduce environmental impact through power-conserving device design and robust power management software, employing low-power processors and wireless connectivity to guarantee reliable performance with minimal energy use. Having remained prominent for some time, green technology usage shows no signs of declining. Grand View Research forecasts that the green technology and sustainability market size will reach $80 billion by 2030, growing at a CAGR of 23% from 2025 to 2030.

By processing data locally, edge analytics software reduces transfers to the cloud, which allows for saving energy and cutting carbon emissions alongside low-heat hardware and micro data centers powered by renewable energy. Low-power chipsets and energy-efficient communication protocols (LoRa or BLE) minimize power consumption, extending device lifespans and reducing battery waste. Solar-powered and other energy-harvesting sensors enable battery-less operations, decreasing maintenance costs and ecological footprints.

IoT-based systems are also widely used to support sustainability initiatives across various industries and application areas, including precision agriculture to optimize water and fertilizer use, smart grids to enable demand-response for balanced energy distribution, refrigerators in supermarkets to optimize cooling cycles and transportation systems to reduce emissions through smart routing and fleet management.

Sustainability-driven IoT principles can also be applied in other IoT contexts, such as industrial automation, manufacturing, and healthcare.

4. IoT-based digital twins

According to Research and Markets, the digital twins market will reach $154 billion by 2030. Digital twins combine IoT data, edge analytics, and AI to optimize decisions in a virtual environment before executing them in the real world. Virtual models of physical objects can be used to predict equipment behavior as well as forecast failure and safety risks, while digital twins of an organization (DTO) help in planning enterprise-level operations and workflows.

5. Advanced connectivity

In 2026, companies are prioritizing next-generation connectivity technologies to enable uninterrupted data flow across IoT networks, which is critical for devices positioned across multiple locations.

5G

5G brings fast speeds and large network capacity, making it possible to process data instantly and power IoT systems at scale.

Wi-Fi 7

With its expanded bandwidth, Wi-Fi 7 permits 320 MHz channels, which are ideal for bandwidth-heavy data transfer, such as 4K video streaming.

LPWAN

Designed specifically for IoT, this technology enables long-range communication with minimal bandwidth and energy use, being cost-effective for managing large numbers of connected devices, such as in utility monitoring.

Satellite connectivity

Satellite networks enable global asset tracking and connectivity for devices located in isolated regions where terrestrial networks fail. GPS data is sent from the device to the central hub immediately, so in case of an emergency, the issue can be addressed in time.

In conclusion

For IoT solutions to support complex operational processes and provide data-driven insights, high-performance software and hardware are critical. Therefore, in 2026, the focus is on creating smarter, more resilient, reliable, and easier-to-manage IoT systems. New technologies are helpful for that by enhancing data analysis, establishing stronger connectivity, reducing operational failures, and improving the use of resources.

To build secure and sustainable IoT solutions, it is beneficial to follow the latest trends to keep pace with technology advancements while setting the standard for resilience and growth.

Even as I've tracked every twist and turn in the technology for autonomous vehicles for going on 15 years now, the key development I've been waiting for occurred, not in the lab or on the road, but in a recent op-ed in the New York Times. In it, a neurosurgeon made the case for AVs as a "public health breakthrough."

He said he was horrified by the more than 39,000 deaths from motor vehicles just in the U.S. last year, "more than homicide, plane crashes and natural disasters combined.... These crashes are also the leading cause of spinal cord injury. We surgeons see the aftermath of the 10,000 crash victims who come to emergency rooms every day. The combined economic and quality-of-life toll exceeds $1 trillion annually, more than the entire U.S. military or Medicare budget."

Then he made the sort of case technologists have been making for years about the potential for AVs to drastically reduce death, injuries and property damage... but this time it came from a doctor. 

And he framed his case as a public health issue, the sort of effort government gets behind and citizens appreciate, even if they may discount the claims they hear from those they see as techno-optimists and rapacious capitalists.

Based on other favorable press in recent weeks and on the relentless rollouts of robotaxis planned for this year, I think we're seeing a sea change. Arthur C. Clarke famously wrote that "any sufficiently advanced technology is indistinguishable from magic," but, over time, the magic wears off, and wildly advanced technologies begin to seem almost normal.

AVs are now being normalized to the point where I think the clock has started ticking on what will be a fundamental rewiring of the auto--and auto insurance--landscape. 

While I've long been an enthusiast about driverless technology, I've always been worried about winning over public sentiment. The techies initially argued that they just needed to be demonstrably better than human drivers and felt that was a fairly low bar, given that more than 100 people a day die in car crashes just in the U.S. But that's not how people look at technology. 

When a driver causes a crash, we may be understanding. We've all done careless things and had near-misses. But if software causes a crash, it wasn't making a mistake in the heat of the moment. Someone designed that software, and they screwed up. The huge company employing that coder somehow missed that mistake, too. 

Machines aren't supposed to make mistakes. Ever. So the bar for AVs' safety is actually higher than it is for human drivers.

Uber stopped its efforts to develop autonomous vehicles after a single fatal accident in 2018 caused a PR disaster. GM's Cruise halted its robotaxi development after one of its cars hit a pedestrian in 2023. The collision was a freak accident, in which another car hit a jaywalking pedestrian and flipped her in front of the robotaxi. But Cruise's AV, programmed to get out of the way after an accident, pulled off to the side of the road--not understanding that the pedestrian was caught underneath the car. The multibillion-dollar AV program couldn't survive the bad PR and scrutiny that followed.

The Cruise debacle left a bad taste. Subsequent press played up resentment of AVs, such as by people who learned they could paralyze a driverless car by putting an orange traffic cone on the hood.

But the press has gradually been shifting. Recently, for instance, Tesla got some nice publicity because one of its cars drove the famous Cannonball Run route between Los Angeles and New York entirely in self-driving mode. (CEO Elon Musk had promised one of his cars would do so by the end of 2017, but still....) Tesla got more attention when Lemonade said it would offer steep discounts to Tesla drivers for miles they traveled while in so-called Full Self-Driving mode, based on the belief that Tesla's AI is much safer than human drivers are, at least in certain circumstances. 

Public opinion has seemed to shift, too, both based on the press and on the growing familiarity with the cars. Yahoo! Finance reports that a survey "conducted in San Francisco this past July found that ​​67% of San Francisco residents now support the operation of driverless robotaxis, up from 44% in 2023, with 'net favorability' of robotaxis swinging from -7% in late 2023 to +38% in mid-2025."

The New York Times piece by the neurosurgeon pulls all those threads together for me and suggests that the public is ready to accept whatever the technologists can deliver. 

Yes, there is always a danger that a car will do something catastrophic. And we'll still see the occasional story about an embarrassing glitch, such as the recent one where a power outage in San Francisco knocked out all the traffic lights, and Waymo's cars just stopped, citywide, because they didn't know what to do.

But, barring a disaster, people are certainly going to see a lot more robotaxis on the road this year. Google's Waymo is already up to about 250,000 paid, fully autonomous rides a week and aims to quadruple that by the end of the year. Waymo already has fleets in San Francisco, Phoenix, Los Angeles, Austin, and Atlanta and plans to add 20 markets this year--including Miami, Dallas, Houston, San Antonio, Orlando, Las Vegas, San Diego, Detroit, Washington, D.C., Baltimore, Philadelphia, Pittsburgh, and St. Louis. Waymo is testing in New York and plans to test in London soon, too.

Waymo is doubling its production of AVs and expects to build more than 2,000 this year. 

And that's just Waymo. Tesla has big plans to expand this year--though any prediction from Musk must be viewed with skepticism, given that he's been consistently overpromising about AVs for more than a decade. Amazon says it will build 10,000 robotaxis a year starting in 2027. Some smaller companies say they're testing robotaxis in Tokyo and Southeast Asia. A host of Chinese companies have pursued autonomous driving aggressively, though generally at the driver-assist level, and the government recently became more cautious after a gruesome accident. Baidu says it will test in London this year, and Europe is shaping up as a battleground. Most of the players there figure to be Chinese and American companies, but Mercedes just announced an autonomous venture with Nvidia, and Nvidia hopes to provide the technology, including simulators, in similar ventures with other manufacturers. 

Insurance won't feel the effect right away, by any means. The tectonic shift won't happen until individuals start buying driverless cars, and when that happens is anybody's guess. But even robust adoption of robotaxis, which should happen over the next couple of years, could be material. If Waymo is really doing 1 million paid rides a week by the end of this year, that's maybe $1 billion of revenue, if annualized. That revenue would have gone to gig drivers and taxi drivers, who buy traditional insurance, but instead will go to a corporate behemoth that self-insures. 

That shift in revenue will mean maybe the loss of just $100 million of premium for auto insurers (based on my back-of-the-envelope calculation). That's a drop in the bucket in a U.S. market measured in the hundreds of billions of dollars of premiums. But exponentials are crazy things. If Waymo quadruples its size this year, what will it do next year? The year after that? And after that?  It's pretty easy to imagine Waymo having 20X its current presence within a few years. And if Tesla, Amazon, Baidu and other Chinese behemoths can deliver, too....

It's worth watching, especially if the public health argument really gains traction.

Cheers,

Paul

Out Front Ideas with Kimberly and Mark kicks off annually with the "20 Issues to Watch" webinar. While there are certainly more than 20 issues to discuss, the focus is on high-impact matters in risk management and employee benefits that require more attention. These are essential issues for every risk manager, HR manager, and insurance professional to monitor in 2026.

1. Connected Risks

Risk does not happen in a silo, requiring assessment across the business and agile planning. Risk managers and their partners must be more diligent than ever in evaluating overlapping risks, including environmental, technological, and human factors. Most organizations recognize their risks, but few are fully prepared to tackle them.

2. Fraud as a Systemic Risk

The Coalition Against Insurance Fraud estimates that fraud costs the insurance industry over $308 billion per year. It can take the form of unethical physicians performing unnecessary surgeries, medical providers billing for services never rendered, and plaintiffs fabricating or exaggerating injuries. Fraud must be met with meaningful consequences, and the industry must actively identify fraud, share intelligence, and demand prosecution.

3. AI Lessons Learned from Early Adoption

Many early AI adopters initially focused solely on platform deployment, only to learn that success hinges on a clear use case tied to measurable business outcomes and a return on investment. Goals for adoption must be clear and concise. Additional considerations include stakeholder engagement, alignment and execution, data readiness, and governance.

4. Industry Engagement

In-person industry events bring colleagues together to help solve problems, exchange ideas, and learn from one another. Conference attendance has not returned to pre-pandemic levels, and that shift has come at a cost to the collective learning and collaboration that strengthen our industry. Reconsider the value of active industry participation and, if given the opportunity, attend a conference.

5. Healthcare Trends

Access to care remains a critical concern, particularly for rural healthcare entities at risk of closure due to continuing physician shortages. As pandemic-era waivers expire, telehealth opportunities are also ending, as physicians can only treat patients within the states where they are licensed. However, AI continues to drive health technology innovation, offering early diagnostic testing and opportunities for self-guided care. Wearables continue to gain popularity, with more industries deploying them to enhance workplace safety.

6. Insurance Market Pressure Points

Legal system abuse continues to worsen the development of liability claims, keeping commercial auto unprofitable despite a decade of premium increases. In response, there is growing interest in quota-share liability towers and captives. The property market avoided hurricane impacts last year, but those benefits were offset by wildfires, with hail and severe convective storms now driving most global catastrophe losses. Workers' compensation remains competitive, yet deteriorating claims have pushed combined ratios above 100% in California and Nevada, signaling an end to the prolonged soft market and flatter rate expectations ahead.

7. Catastrophe Risk Becomes Baseline Planning

For risk managers, what was once an ad hoc emergency response has become a structured playbook to follow in the event of a catastrophe. Some are even shifting from a coordinated team to a business unit that oversees climate, business continuity, and catastrophes. This approach outlines clear roles, responsibilities, and consistent expectations in the event of an incident.

8. Claims Insights

Medical inflation in workers' compensation has historically lagged behind broader healthcare inflation due to fee schedules, but those pressures are now clearly emerging. The National Council on Compensation Insurance (NCCI) reported a 6% increase in both indemnity and medical claim severity in 2024, while the Workers' Compensation Insurance Rating Bureau (WCIRB) in California noted a 9% increase in medical costs. Additionally, expanded mental health claims, catastrophic injuries, and cancer presumptions for first responders are affecting long-term costs.

9. AI in Business Transformation

Fluency in AI is becoming essential for organizations as they adapt to challenges. When paired with user-centric design, AI can drive transformation by improving efficiency while still relying on employees' critical thinking. Organizations that hesitate to adopt automation risk falling behind, especially as time savings can be reinvested in innovation.

10. California Workers' Compensation

California remains one of the costliest workers' compensation states. Savings from the 2012 reforms have been eroded by rising litigation and medical inflation, driving a 127% combined ratio in 2024 and prompting renewed reform discussions. The primary cost driver is cumulative trauma (CT) claims, which broadly cover degenerative conditions and account for over 21% of claims and 38% of litigated cases. While meaningful reform would require addressing CT claims, political resistance makes significant change unlikely.

11. Employee Benefits

Employers are prioritizing engagement, retention, and culture through continuous employee listening and lifecycle surveys. At the same time, health plan costs continue to rise, with projected increases of 6.5–7.6% in 2026, according to Mercer, and growing concern over GLP-1 drug spending. While point solutions remain popular, complexity is increasing. Employees increasingly value purpose, belonging, well-being, and psychological safety as organizations brace for tighter budgets and slower pay growth.

12. Legal System Abuse and Tort Reform

Between 2023 and 2024, the number of verdicts exceeding $10 million increased by 50%, while verdicts over $100 million surged by 68%. These trends can be exceptionally difficult to reverse. However, there has been incremental success, with Florida and Georgia enacting reforms to curb litigation abuse, and several other states considering similar legislation. For the impact to truly resonate with the public, the focus must shift to how these verdicts affect everyday life, including lost jobs, higher prices, and reduced access to services.

13. Workplace Mental Health and Well-being

Supporting psychological well-being is a strategic imperative for engagement, productivity, and safety. Burnout and mental health directly affect business performance and recovery outcomes. Mental health claims now rank second only to pregnancy in leave and disability, surpassing musculoskeletal injuries, and are increasingly recognized as barriers to injured worker recovery. Early identification and targeted support, including behavioral health resources and virtual care options, can improve outcomes and shorten recovery timelines.

14. Cyber Risk

Cybersecurity remains a top concern as ransomware attacks scale through ransomware-as-a-service, expanded attack surfaces, and third-party vulnerabilities. Many breaches go undetected for months, and repeat attacks are common when weaknesses persist. AI-driven tactics, including deepfake executive scams, are increasing risk. Human error remains the weakest link, making continuous employee training, phishing simulations, and healthy skepticism essential to an effective cybersecurity strategy.

15. Workforce Considerations

Retaining today's workforce begins with understanding employee expectations around growth and flexibility. Employees increasingly value career mobility, skills that support current roles, and opportunities to build future capabilities. In hybrid environments, organizations are expanding virtual reality training and self-paced, high-impact "burst" learning programs. Clearly defined career paths are more important than ever, particularly as expectations for flexibility rise. PwC's 2025 research found that 58% of employees would rather quit than return to full-time office work, up from 35% in 2023.

16. Public Entity Challenges

Workers' compensation presumptions and heightened law enforcement liability exposures present unique risks for the public entity sector. At the same time, public entity risk managers face constrained budgets, limited staffing, and aging infrastructure. Pension liabilities remain a significant concern, with recent estimates placing nationwide unfunded public pension obligations at $1.2 trillion. As claim costs rise, higher taxes are likely to follow, and when revenues fall short, essential services are reduced. Because taxpayers ultimately bear these costs, these challenges should matter to private sector businesses and risk managers as well.

17. Reputational Risk in a Real-Time World

Reputational risk is a concern throughout organizations, whether a social media post misses the mark or an operational blunder occurs. Reputational events may affect business success, customer relationships, and growth opportunities. When a crisis occurs, preparation is critical. The response often determines the extent of reputational damage and risk exposure. Knowing how to frame that response for the intended audience is essential and requires understanding stakeholder, employee, and customer perceptions in advance.

18. Regulatory Overreach and Unintended Consequences

Overregulation continues to create significant challenges for businesses and risk managers. In-state physician licensing requirements temporarily waived during the pandemic improved access to care, but those requirements have since been rolled back. Similar regulatory friction exists for claims professionals, particularly in the handling of in-state workers' compensation claims. While well-intentioned, these regulations can fail to keep pace with technology, market realities, and evolving risks.

19. Operational Readiness in the Age of AI

As technology reshapes business models, organizations must ask whether their operations are ready for the future. Without adaptation, some business models risk becoming obsolete within the next decade.

20. Critical Digital Infrastructure: Data Centers as a System Risk

None of today's AI-driven innovations are possible without the massive data centers that power AI and cloud-based systems. These facilities have come under increased scrutiny as some communities court them, while others resist due to strain on critical infrastructure, particularly electrical grids and water supplies. Data centers also create substantial downstream risk due to their role as critical service providers. A single outage can disrupt operations for thousands of businesses.

Listen to the archive of our complete Issues to Watch webinar here. Follow Out Front Ideas with Kimberly and Mark on LinkedIn for more information about coming events and webinars.

What if we have been looking at AI from the wrong angle? What if it is not a magic fix for the insurance industry’s legacy issues but is an unlock for the next generation of growth through insurable risks? 

AI is emerging alongside forces already reshaping the global risk fabric: the rise of intangible assets and cyber exposure, mounting climate volatility, shifting global demographics, and an entirely new class of technologies. These are not distant scenarios, they are today’s realities. 

The IIS Innovation Report reflects an industry in transition, a theme underscored during our executive working group session at the Swiss Re Centre for Global Dialogue in Rüschlikon. Leaders recognized that early AI efforts often focused too narrowly on efficiency and missed the broader strategic opportunity emerging across the global economy. 

The discussions made it clear that the next decade will divide the sector between organizations making marginal improvements and those rebuilding their operating models around proprietary knowledge graphs, reengineered data flows, and augmented human judgment. 

These foundations enable stronger risk selection, superior service performance, and loss prevention in a far more dynamic risk environment, while preserving what remains fundamentally human in our business: trust, advice, and long-term client relationships. AI agents can deliver transformative gains, but only for firms prepared to rethink governance, decision rights, talent, and data strategy. 

This is the strategic inflection point. If we mobilize for it, insurance will not simply adapt, it will become one of the defining stabilizers of an increasingly connected and AI-enabled world!

--George Kesselman

Executive Summary

AI transformation is sweeping the insurance industry

The IIS Report on Innovation, which draws from a diverse respondent pool across insurers, reinsurers, insurtechs, and consultancies, finds that, while enthusiasm for AI is high, maturity levels vary significantly by company size and type. Larger firms are generally further along in production deployment, while smaller firms are focusing more on exploration and customer-facing innovation. 

Efficiency remains the primary driver of AI adoption 

Operational efficiency and workflow optimization dominate current AI priorities, with 53% of respondents citing them as top focus areas, followed closely by underwriting, pricing, and claims management. These findings indicate that insurers are initially using AI to strengthen core processes rather than disrupt existing models. Smaller firms, however, show a stronger tendency toward leveraging AI for customer service and market expansion. Metrics of success largely center on productivity gains, data accuracy, and improved customer experience, though formal frameworks for ROI measurement are still evolving across the industry. 

Experimentation is widespread but deployment maturity is limited

Adoption data reveal that about 87% of companies are pursuing GenAI initiatives, though only around a quarter have reached production-level implementation. Budgets dedicated to AI average 3.9% of overall spending. Most firms rely on third-party general-purpose large language models like ChatGPT, while larger organizations increasingly explore first-party or industry-specific models. Leadership of AI innovation typically originates at the executive level – especially CEOs, boards, and CTOs/CIOs – indicating strong top-down strategic ownership of AI adoption.

Key challenges focus on governance, data, and talent

This report also identifies major challenges that can temper progress. Chief among these are concerns over data privacy and integrity, security, and bias management, as well as the difficulty of measuring ROI. Talent shortages and the lack of formal governance frameworks also impede scalable AI integration, especially for small firms. Most companies rely on human oversight rather than structured governance systems, though larger insurers are beginning to formalize processes through ethics committees, audit trails, and explainability standards.

Innovation is balanced with risk in the era of AI agents 

Looking forward, the report highlights both excitement and caution surrounding the rise of autonomous AI agents in insurance. Top concerns – such as hallucinations, validation difficulties, and regulatory compliance – reflect an industry still grappling with trust and accountability in automated decision-making. Overall, the findings portray a sector experimenting, learning, and building the foundations for responsible, scalable AI adoption that enhances both operational excellence and customer experience

To download the full report, click here.