A diversified book of business can help agents build more resilient, prosperous operations, especially as they trend toward leveraging niches and competition intensifies. For as many agents who are entering into niches, few are doing it effectively. Niche markets are not about agents filling a gap; they must understand where their agency can drive growth, build credibility and prove value to clients.

What are the trends?

Too often, agents jump into a new market without completely committing and lacking a clear plan. Though competition is rising, those agents who dedicate themselves to being a student of their niche and seeking the people and resources needed to help them develop their offerings will be better positioned for success.

Technology can help agents get up to speed on a particular niche, but it is important they remain careful about information they receive and put into artificial intelligence (AI) tools like Google Gemini or OpenAI's ChatGPT. All information received from these large language models (LLM) should be fact-checked as AI can misinterpret information or output false information. Agents must also remember that any information they put into such tools, including client information, can be used by the tool's development company to improve them. They could inadvertently violate a client's privacy or put their own agency's data at risk by adding it to an LLM tool.

While there is no such thing as a perfect starter niche for agents, they should be aware of emerging niches such as cyber, special events, sustainable or green markets and technology companies. Opportunities in emerging markets continue to grow as consumers become more aware of their risks and needs.

Make a choice and follow through

Not every niche will be the right fit for an agency. Agents need to take the time to identify the best choice for their agency, including resources, and then plan to invest in and grow their niches. Consider the following:

• Find the best fit: Agents should start by finding a niche that interests them or connects to a passion. Doing so will help agents commit to developing the niche. They should then ensure there is enough opportunity within their chosen niche. For example, if an agent is passionate about boating but operates hundreds of miles from the nearest marina, it may not be a viable option.
• Do the homework: Succeeding in a niche goes far beyond understanding basic coverages. Agents should work to understand the industry they are getting into, how it operates, related regulations, the client base, challenges, risk factors and more. Having this initial knowledge of a niche will give agents a great jumping off point to provide value to clients, while still learning along the way.
• Build a community: To champion a niche, agents will have to dive into the niche's community and build a network of colleagues and clients. This requires more than simply joining associations and industry groups. Agents need to get involved. Whether in the form of serving on committees, attending meetings, speaking at events or sponsoring relevant initiatives, agents can build credibility by being visible and active. Beyond credibility, industry organizations like SIAA can connect agents to carrier partners who understand and have the ability to support agents' needs within a niche through knowledgeable underwriters. A group like SIAA also provides access to local growth coaches for additional assistance.

A strategy in practice

An agency needs to be prepared to give the time and resources necessary to develop a niche. Agents should have a strategy for how they will implement the niche into their operations, create a team of employees with defined roles who will work on it and communicate a clear plan including how it will be marketed, communicated and benchmarked. It can take anywhere from three to six months for agencies to fully implement a new niche. Agents should take their time through the process, accumulate the necessary background knowledge, invest in their networks and develop a plan to feasibly service and grow a new market.

Niche markets could be an agent's key to developing their agency and maintaining strength in an increasingly competitive market. But those agents who go beyond the surface and become trusted advisors in their niches will benefit from stronger relationships, happier clients and more resilient businesses.

Observability transforms investments from opaque expenses into measurable value drivers in the enterprise. Beyond monitoring, it provides transparency into system behavior and business impact. Leaders who implement comprehensive observability across data, models, LLMs and agents gain competitive advantage through improved outcomes and demonstrable ROI.

In today's dynamic business environment, organizations are making unprecedented investments in data, AI and machine learning capabilities. Yet leaders face a common challenge – What is the business value of my data and AI investments? How do I measure its impact and ROI? 

This scenario plays out repeatedly across many organizations. It is not just about implementing technology, but creating transparency and mechanisms for measuring impact and ensuring systems are perpetually adaptable to changing business needs. This is where observability comes into play.

What Is Observability, and Why Is It Important?

Observability isn't just about monitoring technical performance – it's about creating a direct line of sight between data and AI investments and business outcomes. This transparency is crucial for building trust and demonstrating value. While monitoring tells you when something is not working, observability helps you understand why it broke and how to fix it.

Key components of observability are:

• Data Observability – It focuses on data quality, lineage and metadata management. This includes monitoring data freshness, schema changes, distribution shifts, etc. that affect business. For instance, augmented data quality empowers organizations to prevent data quality problems at the point of ingestion (reducing data down time). It takes data quality to the next level by helping organizations understand data health and performance. Tools in this space include Acceldata, Monte Carlo, and Datadog.
• Model Observability – It tracks model performance metrics, drift patterns and explainability factors. This ensures AI models maintain accuracy and reliability over time, while providing explanations for decisions. Tools in this space include WhyLabs and Arize AI.
• LLM Observability – It tracks prompt effectiveness, response quality, hallucination rates and token usage of large language models. This is critical for managing both performance and costs. Tools in this space include LangSmith and helicone.
• Agent Observability – It extends observability to autonomous AI agents, monitoring their decision-making processes, actions taken and overall effectiveness in completing complex tasks with minimal human intervention. Tools in this space include LangSmith, helicone and AgentOps.ai.
• Infrastructure Observability – It ensures that the technical foundation supports data and that AI systems remain resilient, scalable and cost-effective. Tools in this space include Datadog, New Relic and Dynatrace.

Applications in Insurance and Healthcare

Organizations driving business outcomes aligned to changing customer needs map observability metrics to specific value streams. Below are some of the uses in insurance and healthcare.

Claims – It is a critical function in insurance value chain and has direct bearing on customer experience, operational efficiency and regulatory compliance. It provides the ability to track intake (first notice of loss, or FNOL) accuracy (structured vs unstructured data), natural language processing (NLP) sentiment analysis confidence scores on customer interactions, model confidence scores such as fraud and subrogation across claim categories, along with explainability for decision making and its overall impact in operations. This specifically tracks how NLP confidence scores correlate with customer satisfaction metrics, allowing for continuous refinement. Insurers with mature observability at the FNOL stage, will see an overall reduction in claim cycle time.

Underwriting – Consider a scenario where a broker renews or adds commercial properties to a portfolio and looks for a competitive quote from an underwriter. As part of the process, large sets of information (e.g.: schedule of values) get exchanged and risk assessment is carried out to understand the property characteristics and its exposure to CAT risks, to determine the risk profile, and to arrive at premium. If there is a change or anomaly in property characteristics such as roof condition via roof confidence score, mis-classification due to difference in primary modifiers such as ISO construction class, etc., observability will enable tracking the data drifts in risk factors, CAT model accuracy/performance characteristics etc. that affect the underwriting decision and improve overall broker and underwriting experience.

Clinical Diagnostics/Decision – AI agents alleviate the complexity of this data-intensive process by leveraging LLMs to extract context-specific clinical entities from medical records, summarizing medical conditions, invoking tools such as recommendation engines/agents for treatment options based on patients' medical history and guidelines, etc. In the absence of observability, there will be limited visibility on whether models are hallucinating or adhering to the clinical guidelines, on accuracy or consistency, on the rationale for recommended treatment plans. Overall, AI intervention without observability may become a liability. Hence, observability becomes a cornerstone to transform clinical decisions and improve patient outcomes while saving physician time.

The Way forward

For a chief data and analytics officer, chief data officer or business executive, the business case for comprehensive observability has never been clearer. Below are some best practices and next steps, as leaders are envisioning how to leverage the technology to transform their business.

• Ensure your D&A strategy encompasses an audit of your current observability maturity across data, models and infrastructure. Most organizations discover significant gaps that create business risks and limit benefits. Align your observability investments directly to business outcomes rather than technical metrics alone.
• Build observability into your data architecture from the beginning, rather than retrofitting later. This approach typically reduces implementation costs and builds trust with business.
• Develop an observability-driven culture. That enables teams to leverage insights and helps to address issues at the onset, before affecting business operations. This fundamental shift transforms data teams to value drivers.
• Don't make a one-time investment. Develop a continuing capability that adapts along with your data, AI and business initiatives.

In a nutshell, organizations gaining competitive advantage aren't those with huge AI investments but the ones with visibility into how their data and AI systems create business value. By implementing comprehensive observability, you transform data and AI investments into measurable, adaptable capability for sustainable innovation.

Electric bicycles, or e-bikes, have become a common sight these days in just about every urban environment all over the world. They're fast, efficient, and eco-friendly, making them a popular alternative to traditional transportation. But their rise in popularity is accompanied by a surge in serious accidents and injuries, especially in dense metropolitan areas. As e-bike accidents become more frequent and more severe, the question isn't whether we should embrace this mode of transport– it's already a fact of life– but how we can do it safely.

Bike injuries, especially those between electric bicycles and motor vehicles, can have serious consequences. If you've been injured while riding, know your rights, and where to turn for legal support.

The Surge in E-Bike Popularity

More than 1.1 million e-bikes were sold in the U.S. in 2022 alone, a nearly fourfold increase since 2019.

This growth is driven by multiple factors:

• Rising fuel prices and eco-consciousness
• The desire to avoid crowded public transit
• Advancements in battery and motor technology
• Local rebate and micromobility incentive programs

E-bikes are also capable of higher speeds than traditional bikes, often reaching 28 mph. Many newer models include throttle-only features, removing the need for pedaling entirely. Unfortunately, many riders are unaware of the risks these features present, especially in areas lacking proper bike lanes or municipalities not designating areas for their use, or having no infrastructure designed to support micromobility.

As adoption continues to rise, so do accidents. Emergency departments are reporting a steady increase in micromobility injuries, many involving e-bikes. 

E-Bike Accidents on the Rise

The increase in accidents involving e-bikes has raised serious concerns for public safety officials and healthcare providers. According to the Consumer Product Safety Commission (CPSC), injuries increased by 21% between 2021 and 2022 alone. Even more alarming, e-bike crashes tend to result in more severe injuries compared with traditional bicycles.

The reasons? Higher speeds, heavier frames, automated riding features, and lithium-ion battery hazards. Unsafe speeds and rider inexperience are among the leading causes of e-bike crashes, especially in busy urban environments.

Between 2017 and 2022, emergency departments across the U.S. saw a sharp uptick in e-bike-related visits, from just 751 cases to more than 23,000 annually. In New York City, 23 of the 30 bicycle deaths in 2023 involved e-bikes. And in California, the number of e-bike incidents rose more than 18 times over five years.

Children and adolescents are among the most affected. Riders under 18 years of age now represent over one-third of all e-bike injuries, with many suffering from head trauma, lower extremity fractures, and internal injuries. Some experts link this trend to underregulated sales, lack of age restrictions, and the appeal of throttle-based models that require no pedaling or skill development.

Not all e-bike injuries stem from collisions. A growing number are tied to mechanical issues and battery fires. Cheap, unregulated lithium-ion batteries have been linked to hundreds of fire incidents nationwide, resulting in property damage, serious burns, and even fatalities. Improper charging and poor quality components put every e-bike owner at risk, even when the bike isn't in motion.

Who Is At Fault in an E-Bike Accident?

Determining fault in an e-bike accident isn't always clear-cut. Unlike traditional bicycles, e-bikes fall into a legal gray area, somewhere between bicycles and motor vehicles. This ambiguity complicates insurance claims, police reports, and legal accountability, especially in urban areas. The most common e-bike accidents involve collisions with other vehicles, particularly at intersections where drivers may misjudge an e-bike's speed.

As an experienced bicycle accident attorney will tell you, liability often depends on how the e-bike was being used, what class of e-bike was involved, and whether traffic laws were followed. The faster the e-bike, and the more motor-reliant, the more likely it is to be treated like a motor vehicle in the eyes of the law.

Here are a few common at-fault scenarios:

• Motor vehicle drivers: Cars failing to yield in intersections, opening doors into bike lanes, or turning without signaling are frequent culprits. These incidents often happen because drivers underestimate the speed of e-bikes.
• The bike rider: Inexperienced riders may travel too fast for conditions, misjudge braking distances, or make sudden movements in traffic. Speeding and improper turns are two of the top cited violations in e-bike crashes.
• Manufacturers: If a crash is caused by faulty brakes, a stuck throttle, or a battery fire, the bike manufacturer may be held accountable under product liability laws.
• Local governments: Poorly maintained bike lanes, lack of designated areas, or missing signage can shift liability to the municipality responsible for safe infrastructure.

Essential Safety Tips for E-Bike Riders

As e-bikes continue to grow in popularity, staying safe on the road is more important than ever, and a few steps can greatly reduce your chances of being involved in an accident or suffering serious bike injuries.

1. Always Wear a Helmet

It may seem obvious, but helmet use remains one of the most effective ways to prevent common injuries, particularly traumatic head injuries. This is especially important for riders of Class 3 bikes, which can reach speeds of 28 mph.

2. Follow Traffic Laws and Ride Predictably

E-bikes may look like traditional bicycles, but in many ways, they behave more like mopeds. Stick to designated bike lanes whenever possible, signal your turns, obey traffic signals, and avoid weaving through cars. Remember: motorists often misjudge an e-bike's speed, especially at intersections and roundabouts.

3. Perform Regular Maintenance

Check your brakes, tires, chain, and battery before every ride. A stuck throttle, brake failure, or battery defect can turn a simple ride into a dangerous situation. Only use manufacturer-approved components, and never attempt to bypass the bike's speed limiter.

4. Handle Your Battery With Care

Improper charging and cheap aftermarket batteries are the top causes of e-bike fires. Always charge in a cool, dry area, away from flammable materials, and never leave a battery unattended overnight. The CPSC has urged riders to report defective batteries and only purchase certified models that meet national safety standards.

5. Ride Defensively in Urban Environments

Drivers may not see you, especially in their blind spots. Make eye contact with motorists at intersections, avoid riding too close to parked cars (door zones), and be extra cautious around large trucks and buses.

Beyond protecting your physical safety, these tips can also strengthen your legal position if a crash happens. A well-documented history of safe riding habits can make a big difference in your case, should you need to work with a bicycle accident attorney after a crash.

When to Speak With an E-Bike Attorney

Even when you do everything right, accidents still happen. And when they do, the consequences can be life-altering. Beyond simple inconvenience and physical pain, a crash involving an e-bike can disrupt your world completely. Victims may experience debilitating fractures and spinal injuries and may encounter lost wages due to recovery time, along with mounting medical bills.

It can be hard to know what to do in the immediate aftermath of a crash, so the best way to be prepared is to know your rights and the legal options available to you. No matter the cause of your e-bike-related personal injury, a skilled bicycle accident attorney can help hold the right parties accountable and pursue the compensation you deserve.

The independent agency channel is approaching a crossroads. Half of all insurance agents are expected to retire within the next decade. Fewer than a quarter of today's insurance professionals are under 35. Meanwhile, consolidation driven by private equity often prioritizes short-term efficiency over long-term innovation, and direct-to-consumer models continue to chip away at the personal relationships that have long defined this industry.

This moment of disruption feels daunting, but it's also an opportunity to reimagine what's possible.

Redefining the future of insurance

Having spent my career building companies in relationship-driven sectors, I've seen how technology, when applied carefully, can elevate rather than replace human connection. The same potential exists in insurance—if we focus not on automating the agent out of the process but on empowering them to do more of what they do best.

After hundreds of conversations with agency owners across the country, I hear the same theme:

"I got into this business to help people, but most of my day is spent on admin work."

The challenge isn't a lack of tools; it's that the tools don't work together. Agencies today face a fragmented tech landscape where systems don't talk to each other, and humans are forced to bridge the gaps.

That's not a sustainable path forward. But there's a better one.

Where AI fits in: clearing the path, not replacing the people

Much of the discourse around AI in insurance focuses on splashy use cases: chatbots, robo-underwriting, algorithmic pricing. But the most transformative opportunities are often the least visible.

The real promise of AI lies in the back office—in removing the repetitive, manual work that prevents agents from spending time with clients. It's not about replacing humans; it's about clearing the way for them to do more of what matters.

This starts with solving the real technology problem facing agencies: connectivity.

Consider just a few high-impact applications of AI:

• System integration: Seamlessly connecting CRMs, AMSs, carrier portals, and quoting tools to eliminate manual data entry.
• Document processing: Extracting structured data from unstructured files like PDFs, emails, and ACORD forms.
• Renewal automation: Orchestrating renewal workflows to ensure timeliness and reduce follow-up fatigue.
• Data validation: Comparing inputs across systems to catch discrepancies before they create downstream issues.

And the great news is, there are a number of VC-backed solutions already solving some aspect of this broader agency solution.

Why it matters: Relationships drive results

The financial benefits of automation are undeniable. Many traditional independent agencies are stuck operating at sub-20% EBITDA margins, largely due to the inefficiencies of a disconnected system. While this affects profitability, the greater impact is on the client experience and the relationships that truly drive this industry forward.

Instead of reinvesting hard-earned commissions into fostering and growing these relationships, agencies often find those resources drained by tedious, repetitive tasks. This leaves agents reactive rather than proactive, limiting their ability to build trust, offer meaningful advice, and deepen connections with clients.

Technology doesn't have to undermine the human element of this work. When implemented thoughtfully, it enhances it, allowing agents to step away from the paperwork and focus on what really matters: creating and strengthening personal connections in an increasingly impersonal world.

Done right, automation doesn't replace humanity. It gives people the time and space to truly be human.

A road map for agencies ready to begin

For independent agencies looking to take the first step, I recommend a measured, practical approach:

1. Audit your workflows: Track where your team spends time over a week. Identify high-frequency, low-value tasks.
2. Evaluate your systems: Map out your tech stack and highlight where manual handoffs occur.
3. Start small: Automate one process first. Document handling is often a low-risk, high-reward place to begin.
4. Focus on outcomes: Don't pursue AI for the sake of AI. Define success as time reclaimed for relationship-building.
5. Choose your partners carefully: Whether you build, buy, or partner, connectivity should be the priority, not just shiny features.

The balance of success

My fundamental belief about the future of insurance is that success will come from equal parts innovation and relationships, technology and tradition.

The agencies that thrive will not be the ones that resist technology, nor the ones that abandon the human touch. They will be the ones that use automation to enhance what makes them unique, removing friction so relationships can grow without unnecessary barriers.

In a world that often feels divided between high-tech efficiency and high-touch service, the most resilient path is not choosing one over the other. It is finding a way to bring both together in a way that strengthens each.

The independent agency model has endured for generations because of the value of trusted, personal advice. With the right approach to AI, that model will not only survive the digital transformation. It will help define the future of the industry.

Behind every false claim is a carefully crafted narrative to deceive insurance companies. Fabricated accidents, exaggerated injuries, and staged losses are all framed within seemingly real paperwork. Not only is this an issue for insurers, but policyholders equally bear the brunt of paying for the losses through higher premiums.

An estimated 20% of insurance claims are fraudulent. Even more surprising is that an estimated $308.6 billion is lost to insurance fraud annually in the U.S. alone, costing an estimated $900 per policyholder. Imagine the losses insurers and policyholders bear in the rest of the world. The figure will certainly be jaw-dropping!

But what leads to such a high rate of insurance fraud? Scammers take advantage of the fundamental limitations of traditional fraud detection systems, which is their inability to understand human language. Thus, this has become the new focus in insurance fraud prevention, where the line between genuine and false claims often lies in the subtle details of written text.

There's an urgent need for insurance policy administration systems augmented with natural language processing (NLP) capabilities. These advanced systems can analyze huge amounts of unstructured data quickly and accurately. The result? More effective fraud detection and prevention without compromising the legitimate customer experience.

Before diving into the powerful capabilities and benefits of NLP-powered policy administration systems, it is important to understand why traditional systems fail today. So, let's get started.

Why Traditional Policy Administration Systems Fall Short

Traditional fraud detection systems have rule-based, static algorithms that struggle to detect advanced fraud strategies today. These outdated systems create significant vulnerabilities within insurance operations, leading to serious compliance issues and substantial financial losses. Take a closer look at the limitations of traditional systems:

1. Limited Fraud Detection Abilities

The legacy policy administration system in insurance has limited fraud detection capabilities and struggles to identify subtle patterns that might indicate fraudulent behavior. Moreover, the manual review process creates unnecessary bottlenecks and delays claim settlements. When suspicious activities trigger alerts, investigators often have to dig through documents, slowing the entire claims process for all policyholders.

2. Inability to Process Unstructured Data

Important information often lies in unstructured formats, such as adjuster notes, medical reports, and customer communication. Traditional solutions can't interpret these descriptions, given their inability to extract and process unstructured data. This creates significant blind spots in fraud detection efforts, as important red flags remain hidden within claim details, medical notes, and policyholder communication.

3. Reactive Rather Than Proactive

Legacy policy administration systems identify fraud during post-payment audits. More concerning is that these systems aren't connected to data sources; thus, they fail to capture valuable context that could reveal suspicious patterns. This reactive approach drains substantial funds way before fraud is even detected. Besides, recovery efforts prove costly and often unsuccessful compared to prevention strategies.

4. Inflexible Rules and Parameters

Traditional policy administration solutions require manual reprogramming to adapt to modern fraud techniques. Insurers require technical expertise for this process, which creates operational delays and downtime. Fraudsters know how to exploit the static thresholds built into traditional systems.

Not only this, but fixed threshold parameters trigger excessive false alarms that drain investigative resources, failing to identify advanced fraud attempts that circumvent predefined rules. The rigidity of these systems prevents the continuous improvement necessary to fight evolving fraud strategies.

As is evident, insurers direly need advanced and innovative solutions to combat fraud. Integrating NLP in insurance policy administration systems precisely solves the problem. This combination creates a powerful tool that effectively addresses the shortcomings of traditional insurance fraud detection systems. So, let's explore in detail how NLP-augmented policy administration solutions empower insurers to fight fraud.

Role of NLP in Combatting Fraud

Natural language processing lets algorithms understand and interpret natural language as humans do. Thus, it's no wonder that the NLP market size is expected to exceed $200 billion by 2031, growing at a CAGR of 25%. Now imagine adding these capabilities to traditional policy administration systems! These transform fraud detection through advanced linguistic analysis capabilities. Here's what all NLP can help with:

Enhanced Data Extraction and Analysis

Unlike traditional methods, an NLP-powered policy administration system in insurance can extract information from previously inaccessible unstructured data such as claim descriptions, medical reports, adjuster comments, etc. These systems identify inconsistencies between claim narratives and other documentation, revealing fraudsters' attempts at misrepresentation.

For instance, NLP tools can detect when injury details in claim forms contradict accompanying medical reports. These minute but significant contradictions often indicate fraud that might otherwise go unnoticed.

Sentiment and Linguistic Pattern Recognition

Sentiment analysis helps detect emotional clues that might point to fraud. For instance, signs like unusual language patterns, excessive details, overuse of qualifying statements, and passive voice trigger automatic alerts for further investigation. Additionally, emotional distancing language and unnecessary explanations in claim documentation also suggest fraudulent intent. Sentiment analysis is the best way to distinguish legitimate submissions from false ones.

NLP-based policy administration software can identify specific words and phrases often linked to fake claims. These language markers generally appear in multiple fraud attempts in similar situations. The advanced systems analyze semantic patterns that correlate with proven fraud cases. This builds an ever-growing knowledge base of linguistic signs associated with deceptive practices.

Entity and Relationship Mapping

There are instances when insurance fraud is coordinated and organized by multiple entities. Individual claim analysis might not be able to identify such a relationship network across seemingly unrelated claims. Thanks to the advanced capabilities of NLP-based systems, insurers can easily identify and verify relationships among claimants, witnesses, medical providers, and repair facilities.

Additionally, NLP systems excel at the temporal analysis of claim stories. They can detect when the order of events doesn't make sense or when timelines seem unlikely in descriptions of claimed incidents. These timing anomalies often reveal fake scenarios created to support false claims. The system identifies when described sequences have impossible logic or timing contradictions that human reviewers might miss.

Multilingual Processing and Risk Classification

In the context of international scams that usually breed from language barriers, NLP-enhanced insurance policy administration software stands out as an armored knight. It can effectively process multilingual communications with consistent accuracy, ensuring that detection capabilities remain effective, regardless of the document language. Additionally, claims receive risk scores based on comprehensive linguistic analysis. Thus, insurers can prioritize the most doubtful submissions and better focus investigative resources on high-risk cases.

Continuous Learning and Adaptation

What differentiates NLP-powered insurance policy administration solutions from traditional rule-based methods is their ability to learn and evolve. They have cross-referencing capabilities for automatic comparison between current claims and historical data. The system identifies patterns across seemingly unrelated claims that share suspicious characteristics. This comparative analysis reveals subtle similarities that might indicate coordinated fraud attempts or repeated tactics from known offenders.

The systems continuously learn from adjuster decisions and outcomes and improve through exposure to adjudicated cases. This creates increasingly accurate fraud detection models without requiring manual rule updates, thus ensuring detection capabilities remain current against evolving fraud techniques.

Final Thoughts

As insurance fraud rises, adding NLP to insurance policy administration software is not a fancy move but a smarter way to combat scams. In addition to identifying fraud, these advanced systems help boost efficiency and improve customer experiences while preventing financial losses and reputational damage. Even though shifting from traditional systems to new ones is challenging, the ROI quickly materializes through significant drops in fraud.

Remember flip phones? They could make calls, send texts (if you were patient) and maybe took a blurry photo or two. Back then, they felt like cutting-edge tech, and for a while, they got the job done. But imagine trading your iPhone for one of those old Nokias today. Good luck ordering an Uber, paying for coffee or getting to your next meeting without a printed paper map.

That's exactly what it's like when insurance companies cling to outdated legacy systems.

Sure, those platforms might still "work," in the most technical sense, but they weren't built for today's speed, scale or customer expectations. Just like a flip phone in a smartphone world, legacy systems slow you down, frustrate your users, and make it harder to keep up, let alone get ahead.

Outdated Tech, Real-World Consequences

And that's not a hypothetical problem; 74% of insurers are still running on legacy systems, and the consequences are stacking up fast.

Legacy systems are one of the leading causes of customer dissatisfaction in the insurance industry today. These systems slow insurers' ability to deliver the kind of seamless, digital-first experience customers expect, especially younger, tech-savvy ones who compare every interaction against Amazon or Apple.

And the risks go way beyond frustration. In 2021, CNA Financial was hit with a ransomware attack that shut down systems, exposed sensitive customer data, and reportedly cost them $40 million in ransom. The incident made national headlines; customers were rattled, and the brand's reputation took a hit (one it's still working to recover from). So, in today's climate, legacy systems aren't just inefficient, they are a liability. In an industry built on trust, you can't afford that kind of failure.

The damage isn't just visible. There's also the money you're bleeding without even realizing it. Research shows 5–10% of premiums are lost every year to "premium leakage." Translation: That's revenue slipping through the cracks when legacy processes and disconnected or fragmented systems fail to accurately assess or price risk.

Legacy systems may still "work," but they're quietly (or not so quietly) slowing you down, eroding trust, and draining profits.

Modernizing Doesn't Mean Breaking the Bank

Upgrading technology makes people nervous, and honestly, who can blame them? Upgrades bring up all the right, but tough to answer, questions: How much will this cost? Why do we need it now? How long is this going to take?

Those concerns are valid. But here's the catch: Sticking with outdated systems creates bigger, more expensive problems. And the industry knows it; 71% of insurance executives are frustrated by how hard it is to launch digital programs. That's not just noise; it's a red flag. The industry is ready for change, and customers already expect it.

The good news is you don't have to blow up your budget or your team's bandwidth to modernize. Today's platforms are leaner, faster and designed to make life easier for your people and your policyholders.

If you are still on the fence, know insurers using modern policy admin systems have slashed IT costs per policy by up to 41%, according to McKinsey. That's real money you can redirect into innovation, training and delivering the kind of experience that earns trust and keeps it.

Taking the First Step

Modernizing your business doesn't have to mean burning everything down and starting from scratch. In fact, the smartest companies start small, zeroing in on high-impact areas like claims or customer service, and roll out modern solutions in phases. It's a faster path to real results without throwing your operations into chaos.

Legacy systems might've done the job once, just like that old flip phone you couldn't live without. But today, they're more liability than assets. The longer you hang on, the harder it gets to keep up. The real shift isn't just about tech. It's about staying competitive, staying secure and staying relevant. The best advice I can give is to start where it matters, scale toughtfully and stay ahead.

Paul Carroll

There are so many exciting developments in drugs, testing and treatments in healthcare, but there’s little or no historical data about them. How do you provide the information that lets underwriters evaluate risk and develop pricing?

Colin Condie

Data is received from many sources, including insurance carriers, clients, and data vendors. This data is analyzed extensively, with a focus on the frequency and severity of claims and their underlying conditions and prescription drug histories. The predictive model algorithms examine condition categories and prescriptions that occurred in the past and their correlations with claims. Risk scores are then developed based on these relationships. These risk scores are then applied to current conditions and prescriptions of the group’s members to develop expected claims predictions for the group. The risk scores that are developed are then used by underwriters in developing premiums for the groups.

In cases where there is little historical information on drugs to rely on [i.e. a new drug], additional methods such as reliance on clinical data can be used when developing the risk score. The risk scoring models are then updated continuously as new data on drugs becomes available.

Paul Carroll

What are the most significant cell and gene therapy [CGT] developments and FDA approvals that industry professionals should be monitoring for potential market exposure? 

Colin Condie

Currently, cell and gene therapy drugs are primarily targeting three main areas: rare genetic diseases, blood disorders, and certain cancers.

For blood disorders, there are treatments such as Hemgenix for hemophilia B and Roctavian for hemophilia A. For rare genetic disorders, there are treatments such as Skysona for neurodegenerative disorders [i.e. CALD], and Zolgensma for spinal muscular atrophy [SMA]. In oncology, there are treatments such as Kymriah for leukemia and Abecma for relapsed or refractory multiple myeloma.

Another significant area is retinal disease, where Luxturna is a cell gene therapy to treat vision loss that is caused by inherited retinal dystrophy. 

Looking to the future, we're seeing an expansion beyond these traditional areas. New trials are focusing on conditions with greater prevalence than the current rare diseases that typically affect a small segment of the population. For instance, treatments are being developed for refractory angina (chest pain caused by reduced blood flow to the heart) and ischemic stroke (where blood vessels to the brain are constricted).

The future FDA approvals will continue to focus on familiar therapeutic areas while expanding into new areas such as muscular dystrophies and inherited conditions. Neurology is emerging as a new frontier, with gene therapies targeting inherited neurological disorders such as Alzheimer's. 

About 80% of FDA approvals in the pipeline will be for specialty drugs, including cell and gene therapies. These specialty drugs are expected to cost on average between $200,000 and $400,000 annually, representing a continued shift toward innovative, high-impact, and high-cost therapies targeting rare diseases and chronic conditions. 

Paul Carroll

How is machine learning technology helping underwriters navigate the complexities of healthcare data analysis? 

Colin Condie

Machine learning, predictive modeling, and artificial intelligence are already making significant impacts. These technologies are particularly effective at identifying high-cost conditions and prescription histories and predicting associated costs based on this information via the risk score that is used by the underwriter. 

One key application is analyzing data for "one and done" therapies. Unlike maintenance medications that require continuous administration, CGTs are generally designed as single-dose treatments. The artificial intelligence and actuarial models are used to estimate the expected occurrence and cost of these therapies, and this information is then reflected in the risk score. 

In the early stages of CGTs, occurrence rates are relatively low, and therefore there may be insufficient claims data available to use as the basis to reflect their impact in the risk scores and the resulting prediction of future costs. Therefore, the focus is on identifying condition categories that might indicate where the therapy could be appropriate. Additional variables are analyzed, such as patient age, specific diagnoses, diagnosis codes, disease severity, and drug specific data such as FDA clinical trial information. Indicators of prior unsuccessful treatments are examined, as many cell and gene therapies are typically prescribed after other treatment regimens have failed. 

There are also external factors that need to be adjusted for in the predictive modeling. For instance, some therapies may be ineffective for patients with certain antibodies and therefore an adjustment to the assumed frequency of the therapy may be required in the predictive modeling. 

The main challenge is the limited availability of data. The focus is on gathering as much information as possible to determine the expected prevalence and costs for the CGTs, which includes an analysis of medical conditions. The machine learning tools help in the assimilation of data from different data sources from which accurate predictions can be performed by the models.

Paul Carroll

How much does AI reduce turnaround time in underwriting while maintaining actuarial integrity?

Colin Condie

The concept of automated underwriting and quick turnaround times has been a focus in the industry. The predictive modeling uses AI-based algorithms to generate risk scores that represent the predicted health status or morbidity of the members of a group. Along with other variables, the risk scores help determine expected future claims costs, creating a data-driven foundation for underwriting decisions that optimizes efficiency and accuracy.

With the predictive models, underwriting decisions can be automated for groups that are determined to be very low risk or very high risk based on the risk score that is generated. The predictive model results can also be used as an indicator when underwriter review is necessary. For example, the predictive model can flag cases where underwriter review is necessary, such as for a group that has one high-risk member driving the prediction while the other members of the group have low risk. 

Paul Carroll

How will predictive models incorporate data from wearable devices and remote patient monitoring systems into underwriting processes over the next few years? 

Colin Condie

It's interesting. Risk vendors currently base their predictions using health status [morbidity] risk scores based on historical medical and prescription data. Some risk vendors also incorporate lifestyle-related factors as predictors of risk, such as smoking habits, alcohol consumption, eating habits, exercise patterns, body mass index levels [BMI], and sleep behaviors. Predictive models that use wearable devices or remote patient monitoring [RPM] systems can provide both health status-based and lifestyle-based factors for vendors’ risk scoring models. 

Predictive models that use wearable devices and remote patient monitoring systems analyze continuous streams of biometric data, including heart rate, blood glucose levels, blood pressure, respiratory rate, activity levels, and sleep patterns. Targeted interventions can be implemented for members based on these metrics.

Diabetic members can use AI-driven RPM devices to adjust insulin doses, which can lead to a reduction in hypoglycemic events. Wearables can identify heart rate irregularities, which can lower the risk of having a stroke. AI devices can analyze sleep patterns and heart rate to predict potential anxiety or depression episodes, which can lead to improved mental health treatment outcomes.

Paul Carroll

How accurate and effective are health risk scores in predicting medical costs and outcomes? 

Colin Condie

Health status risk scores are accurate. Vendors that also generate lifestyle risk scores provide additional information in terms of the risk of the group over the longer term.

The accuracy of these AI-generated scores is continually evaluated through real-time studies, comparing predicted costs against actual claims data. The health status risk scores are evaluated by comparing the actual claims experienced with the claims that were predicted based on the risk score. The lifestyle risk scores are generally evaluated over a longer-term time horizon because the impact of lifestyle factors on claim costs generally occurs over a longer period (i.e. the number of years it takes for tobacco use to cause the onset of medical conditions). A challenge involves keeping pace with new drug approvals. Risk-scoring models require constant updating to include new drugs and their corresponding national drug codes [NDC]. Additionally, the models must account for evolving condition and drug cost and utilization patterns and incorporate scenarios where the claims experience is immature or infrequent.

An additional challenge is developing assumptions for CGT treatments that have limited frequency. For example, there haven’t been many CGT occurrences overall, and only a few of the approved therapies are widely used. When medical conditions are relied on to estimate the use of CGTs, a consideration is that the claims data may not have the specificity that is required to identify the ideal conditions for CGT treatment. To address these and other issues, data scientists, actuaries, and engineers analyze multiple data sources, including available medical and drug claims data and clinical drug trial information, to determine the impact on risk scores of future CGT utilization and cost.

Paul Carroll

Thanks, Colin. This is super informative.

About Colin Condie

Colin Condie is a senior healthcare actuary at Verikai, where he leverages his extensive actuarial expertise to enhance the company's risk adjustment solutions. With nearly three decades of experience in the field, Colin specializes in predictive analytics for both fully insured and self-insured markets, serving insurers, MGUs, stop-loss carriers, employer groups, and PEOs. Prior to joining Verikai in September 2024, Colin served as director and actuary at ExtensisHR for over seven years, where he developed a deep understanding of the PEO industry. His career also includes actuarial roles at Aon, AXIS Global Accident and Health and Munich Re, building a diverse background across consulting, insurance, and reinsurance sectors. Colin's expertise lies in data-driven decision-making, risk assessment, experience monitoring and reporting, and pricing models. At Verikai, he collaborates with cross-functional teams including product experts, data scientists and engineers to refine risk adjustment strategies and expand the company's presence in the fully insured and self-insured markets. A Rutgers University graduate with a bachelor’s degree in economics and an MBA in finance, Colin brings a unique blend of analytical skill and business acumen to his role. Based in Marco Island, Florida, he approaches actuarial challenges with both technical precision and a broader business perspective.

As insurers leverage the power of cutting-edge technologies like machine learning, artificial intelligence, and third-party data integrations to improve their data-driven decision-making, they're faced with the critical task of ensuring the security and control of sensitive data, which is essential for maintaining customer trust and regulatory compliance. One promising solution to this challenge is a collaborative intelligence ecosystem where multiple parties come together to form a shared knowledge pool that facilitates secure and collective learning through multi-party computations. Each person’s sensitive data and information remain private and undisclosed.

Munich Re Life North America’s Integrated Analytics team believes this innovative technology has the capacity to unlock significant computing and analytics opportunities within the insurance industry. They investigated its capabilities by partnering with an external vendor to conduct a proof of concept (POC). This article shares the details of our POC experience, which evaluated the performance, accuracy, efficiency, and scalability of the vendor’s multi-party computation platform.

Purpose

A vast amount of data is generated daily and growing exponentially due to technologies like social media, cloud services, the Internet of Things (IoT), and artificial intelligence (AI). While this data holds significant value for businesses, concerns about privacy, trust, and other risks leave much of it inaccessible. This is especially true for industries like insurance, where sensitive personal information is prevalent and privacy is a priority.

However, advancements in privacy-retaining technologies offer a more secure framework for collaborative research. Multi-party computation is a cryptographic technique that allows participants to jointly compute a function on their private inputs without revealing those inputs to each other. Our overarching POC goal was to evaluate its potential promise in reducing challenges around data transfers, barriers to collaboration, and vulnerability to bad actors in driving insurance use cases for Munich Re.

How does the technology work?

We worked with a vendor experienced in developing and deploying privacy-enhancing technologies for commercial use. They have pioneered cryptographic privacy-preserving computation technology for analytics and machine learning. Their product enables sensitive data to be distributed across teams, organizations, and regulated jurisdictions by deploying privacy zones (on-premises or cloud) within the infrastructure and network of each party sharing insights.

Through these privacy zones, a compile function is run on each local network using its own data. The function then sends updates to a central server that aggregates the results across parties, effectively keeping sensitive information private within the respective local networks.

For example, if the goal is to train a global machine learning model using insights from different sources, each source would train a local model on its own data and only send the updated model parameters (e.g., weight) to the central server. In our case, the central server was a part of a cloud-based Software as a Service (SaaS) framework provided by the POC vendor.

This concept is called federated learning, which is a decentralized machine learning approach that leverages multi-party computation principles to allow parties to train a model collaboratively, with each preserving its data locally. Instead of sharing raw data, parties share model updates or encrypted data, allowing for secure computation and aggregation of model parameters.

Method of evaluation

The POC had two phases: Phase I assessed the basic multi-party computation functionality within the vendor’s sandbox environment, while Phase II ran a synthetic two-party computation within Munich Re’s cloud environment. For the latter, we trained a popular advanced machine learning algorithm on the multi-party computation platform. This allowed us to securely leverage insights from our internal historical data, which was hosted in one privacy zone, and enhance it with externally acquired sociodemographic data, which was hosted in another.

We ultimately drew insights from both datasets while ensuring privacy within each computation zone. Our goal was to evaluate the predictive accuracy of the resulting model, lift from increasing samples and features, ease of setup, and efficacy of the privacy-preserving computation process.

Our evaluation was split into four broad categories:

Functionality:

• Preservation of data privacy and security behind each party’s firewall
• Dedicated functions to join private data (private set intersections)
• Granular data privacy controls
• Complex operations available to be implemented for data preparation
• Functionality to run advanced algorithms

Efficiency:

• Speed and quality of technical support
• Speed and efficiency in running functions and algorithms
• Extent of coordination required among parties for data processing and modeling

Correctness:

• Accuracy of results from running statistical functions, algorithms, and data processing operations

Scalability:

• Potential to scale up for faster, optimized multi-thread processing
• Level of data engineering expertise required for initial setup

Overall, our assessment of the technology and its potential is favorable. We believe the integration of innovative techniques like multi-party computation can provide a reliable way to enhance business capabilities, enabling secure and private data sharing and analysis across multiple stakeholders while maintaining data confidentiality and integrity.

Potential use cases

Internally within larger companies: To utilize data across entities/departments to increase data size and features for analytics or model development.

Externally with partners and third-party data vendors: For fast and efficient data evaluation. Raw data is never shared, and sensitive information is protected, mitigating data security and privacy risks.

We believe this technology holds significant potential. It offers an efficient way to meet existing privacy compliance and data sharing best practices in building collaborative intelligence ecosystems within and across organizations.

It's only a matter of time before companies use multi-party computation frameworks to enhance their informational edge.

Sponsored by ITL Partner: Munich Re

My husband and I do our grocery shopping once a week for just the two of us and our five cats—though we don't include cat food in our weekly grocery budget. Typically, we purchase the usual staples: fruits, vegetables, meats, bottled water, soft drinks, and occasionally snack items. By the time we reach the checkout, our total averages more than $180 per trip. Yes, nearly $200 a week—for just two people.

If groceries alone cost us this much, what are households of three or more people paying—and how do they manage to afford it? And this is before adding in other common household expenses, like mortgage or rent, utilities, transportation, debt, internet and other subscriptions, family needs, savings—and finally, insurance.

Inflation pressure is real, and we continue to live it. MarketWatch reported that as of December 2024, "consumer prices were up 2.9% year over year," contributing to a steady rise in the cost of living.

And it's not just groceries and household expenses. We are also seeing our insurance premiums increase year over year.

Effects on Consumers and Business Owners

According to the U.S. Department of the Treasury, average homeowners' insurance premiums per policy increased 8.7% faster than the rate of inflation between 2018 and 2022. More recently, in 2024, the national average cost of homeowners insurance rose to $2,728 per year.

Auto insurance hasn't been spared either. Bankrate reports that the average cost of full coverage car insurance climbed to $2,638 in 2025, up 12% from 2024.

Consumers like us aren't the only ones affected. Commercial businesses are grappling with increased operating costs, pricing pressures, supply chain disruptions, wage inflation, rising financing and marketing expenses, shrinking consumer demand, and uncertainty in long-term financial planning. Like households, businesses also face rising insurance premiums. According to WTW's Commercial Lines Insurance Pricing Survey, U.S. commercial insurance rates increased by 6.1% during the third quarter of 2024. Specific lines, such as commercial auto insurance, have seen even steeper increases, with rates continuing to rise in double digits.

Effects on Insurance Carriers

Insurance carriers, like homeowners and business owners, are equally affected by these economic pressures. Rising repair costs for homes and vehicles are driving up claim payouts. The National Oceanic and Atmospheric Administration (NOAA) reported that damages from weather-related disasters in the U.S. amounted to approximately $92.9 billion in 2023. These factors are forcing insurers to raise property insurance rates.

At the same time, insurers face a raft of other issues—more frequent and severe weather events, rising reinsurance costs, and a surge in lawsuits and settlement amounts. This last factor, social inflation, was cited by Swiss Re as a key driver of higher liability claims costs, particularly in lines of business exposed to bodily injury claims.

Combatting Inflationary Pressure With Agentic AI

To manage these mounting pressures without relying solely on rate increases, insurers are turning to agentic artificial intelligence (AI) solutions. An agentic AI platform provides specialized AI agents that automate routine business processes across various operational domains. These agents operate continuously, delivering high accuracy in tasks such as data extraction, document classification, and workflow orchestration.

This approach equips insurers to combat inflationary pressures through smarter operations, tighter cost control, and enhanced customer service by delivering:

• Operational optimization: Agentic AI automates repetitive tasks such as data entry, document processing, and information verification. This minimizes manual data review, reduces the risk of errors, and accelerates process timelines to increase productivity.
• Cost savings: Boosting operational efficiency can significantly reduce administrative overhead and processing costs. Freed from routine tasks, underwriting and claims experts are able to focus on product innovation, customer engagement, and other higher-value tasks to improve overall ROI.
• Service improvement: Agentic AI enhances service delivery across underwriting and claims management workflows. It automates critical stages such as policy intake, document triage, reserve allocation, and policyholder communications. This results in greater accuracy, reduced cycle times, accelerated claims resolution, providing policyholders with quicker, more responsive service and fewer disputes or delays.

These innovations position insurers to enhance profitability, elevate customer satisfaction, and build a lasting competitive advantage, even in today's demanding economic landscape.

Faced with escalating economic challenges, including inflation that affects every part of the insurance value chain, insurers must turn to AI as a vital tool for resilience and growth.

Financial pressures from rising claims severity, operational inefficiencies, and other challenges highlight the urgent need for transformative solutions. Agentic AI offers a strategic path forward—reducing costs, enhancing service delivery, and helping insurers do more with less. As inflation makes consumers increasingly price-sensitive, AI-driven automation enables insurers to stay resilient, responsive, and competitive. By investing in AI, insurance companies can not only navigate today's economic pressures but also build a more equitable future for policyholders.

The auto insurance industry is experiencing another transformation, led by the preferences of the newest generation of drivers entering the market – Gen Z.

Born between the mid-1990s and early 2010s, this tech-savvy group has grown up in a digital world and demands more from auto insurers than previous generations. They seek seamless online experiences, personalized services, and transparency in pricing and claims processes.

As they continue to become a significant force in the market, insurance companies must adapt their strategies and processes to cater to the unique needs and values of this latest generation of drivers. Here are three key things that Gen Z wants for auto insurers to keep in mind.

1. More transparency in pricing

Gen Z drivers are less likely to shop around for auto insurance than other generations. According to a 2024 report on car insurance shopping trends, only 19% of Gen Z respondents compare auto insurance prices annually, compared with 42% of Millennials.

Hidden fees and complicated pricing structures are major turn-offs for Gen Z. They want clear, upfront pricing that outlines exactly what they are paying for. When choosing an auto insurance policy, Gen Z drivers prioritize coverage options (61%), followed by premium cost (49%), and customer service reputation (34%).

Interestingly, Gen Z also places less emphasis on premium costs compared with older generations, with coverage options being their top priority. For insurers, it's important to have transparent pricing and provide detailed explanations of policies.

2. Digital-first claims experience

To keep Gen Z drivers happy, the claims process also needs to be frictionless and digital-first. This generation of drivers has grown up in a world where digital interactions are the norm, and they expect the same level of digital engagement from their auto insurers.

Allowing drivers to instantly upload damage evidence and necessary documents, track claim status in real-time, and receive clear communication at every step can greatly improve customer satisfaction with this generation. Insurers can upgrade any outdated claims processes through the use of artificial intelligence (AI) technologies like visual intelligence. An advanced type of computer vision AI, visual intelligence can accelerate new policy subscriptions, claims assessments and cash settlements. The technology also enables policyholders to submit evidence remotely at the incident scene so insurers can analyze and potentially resolve a claim within minutes of first notice of loss.

Auto insurers are increasingly adopting AI technology to guide younger policyholders through a remote, yet thorough, evidence-gathering process. It is also helping auto insurers and body shops provide pre-repair cost estimations faster, decreasing the need for physical vehicle inspections. Once evidence is collected remotely, visual intelligence allows insurers to compare it immediately against a database of previously gathered evidence, identifying unusual cases and flagging them for human intervention. This not only speeds up the claims process but also reduces the risk of fraud.

A smooth mobile experience enables young drivers to manage their insurance policies and submit claims anytime and anywhere they need it. Furthermore, whether it's through chatbots, live chats, or social media, AI-powered customer support helps insurers meet Gen Z's demand for instant responses, freeing humans to focus on more complex cases.

3. Personalization in services and support

The era of a single policy that meets the varied needs of all drivers is over. Gen Z wants policies that reflect their individual driving habits and lifestyles. Usage-based insurance (UBI) models, which adjust premiums based on actual driving behavior, resonate well with this demographic. By using telematics data, insurers can offer personalized policies that reward safe driving and give better value for money.

Gen Z drivers also show a strong preference for embedded insurance options. 79% of Gen Z drivers would prefer having insurance integrated into the car deal itself, and 81% stated they would like the option to purchase insurance at the point of buying their vehicle.

To connect with Gen Z drivers, automated emails and generic customer support messages also won't cut it. Customer communication needs to feel personal. AI-driven tools can help insurers create personalized strategies; whether it's promotions or policy updates, every interaction can feel unique and relevant. AI algorithms can analyze customer data, driving habits, and social media activity to create messages that resonate with Gen Z drivers' interests and needs.

The next generation of auto insurance

By adopting a digital-first strategy, offering personalized services, and being more transparent, insurers can build strong connections with the latest generation of drivers. Using advanced technologies like AI can help insurers enhance the customer experience and stay relevant in the market.