It’s the rare organization these days that has not at least begun to contemplate using generative AI (GenAI) to unlock new opportunities and enhance workflows. There is good reason for this excitement; GenAI represents game-changing possibilities across industries, including insurance. But, like any new technology, it also comes with potential pitfalls that organizations are well-advised to address before proceeding.

Admittedly, it’s hard not to be enthusiastic about many of the potential benefits of GenAI, including its ability to elevate customer experiences, refine performance in certain areas and drive operational efficiencies. But wise executives will nonetheless proceed with caution when it comes to integrating these new tools into their technology ecosystem. 

If your insurance organization is considering GenAI integration, here are four key questions to ask yourself first.

Does this technology align with your organization’s ethical, security, data and client consent standards? 

Implementing trustworthy GenAI is paramount for business optimization, improved outcomes and reputation protection. Before adopting the technology for any business use case, it's essential to identify and address potential concerns tied to ethics, culture, human factors or change management. 

Once you have confirmed alignment with your organization’s standards, proper training and comprehensive risk assessment are crucial. To bolster trust and responsible AI deployment within your business, it's imperative that GenAI – like any new technology – undergoes thorough security and data privacy vetting to help ensure its ethical use and strategic incorporation. 

Implementing AI into core business functions also demands rigorous testing and validation. Take the challenge of bias in an image detection AI model. When AI is trained on skewed data, it can produce biased performance, which in turn can distort applications that end up relying excessively on these biased sources. To mitigate this problem, companies should diversify and balance their training data, adopting continuous monitoring and retraining practices and seeking diverse stakeholder input during regular audits. 

How will this tool lift up operational, customer and employee experiences?

Executives across the insurance industry are eager to see how GenAI tools can improve processes and increase efficiencies across enterprises. But as tempting as it may be to forge ahead with the technology, the integration of GenAI should always begin with the assessment of current processes. 

Insurance companies should consider GenAI as a tool to enhance processes, not replace them. While technology empowers the business, optimizing workflows is crucial – after all, existing flaws might be amplified by integrating technologies like GenAI. Identifying the pain points in existing workflows can help businesses understand where technology implementation is needed and what technologies can reduce, or in some cases remove, a problem.

See also: 3 Key Uses for Generative AI

In the backdrop of varied legal landscapes, are you adhering to regional laws? 

Given the evolving nature of data privacy laws and the high stakes surrounding client consent, organizations must approach GenAI with utmost caution. A misstep could result in significant legal and reputational repercussions.

Like all evolving technologies, the innovation that GenAI brings comes with its own set of concerns. Large language models produce human-like content but face challenges like misinformation, malicious use and opaque decisions. Unlike deterministic systems that predictably respond to set rules, GenAI operates probabilistically. As a result, it sometimes generates content that is disconnected from reality. In addition, its immense scale presents potential issues with interpretability, bias and control. 

A core concern is AI's potential for delivering results that, while appearing structurally sound, may not always be factually accurate. It's important to note that while AI systems can process vast amounts of information swiftly, their outputs require rigorous validation. Insurance companies should consider implementing a robust AI governance framework so they can harness AI's potential while ensuring that the trust and reliability that their clients expect remain uncompromised.

While deterministic AI provides consistent results and probabilistic AI embraces uncertainties, neither can fully capture the nuances of human understanding. This is where the "human in the loop" approach comes in. Humans bring empathy, ethics and contextual understanding to the assessment of AI outcomes that machines may overlook. This collaboration ensures that AI technology serves as a complementary tool, fostering decisions that are balanced, fair and contextually relevant.

See also: 5 Ways Generative AI Will Transform Claims

Is your GenAI investment calibrated for optimal ROI, and does it fit with your growth goals?

With a more complete understanding of when and how to use GenAI effectively and responsibly, your company will be in a strong position to enhance efficiency and effectiveness through the integration of this technology. Careful consideration of your organization’s goals and objectives and regular assessment to ensure tangible ROI will round out the successful planning and evaluation process for your GenAI investment. 

Remember, for effective AI integration, it's vital to test consistently, prioritize explainability and maintain robust performance to quickly address potential model or data discrepancies. It's also crucial to establish governance policies for assessment and strategy, culminating in a responsible AI framework solution.

With all of these pieces in place, your organization will be in a strong position to unleash the power of GenAI. 

Then -- a history lesson

Raise your hand if you remember when agents faxing hand-written applications to underwriters was commonplace. It wasn’t that long ago, was it?

Now that we have you yearning for the days of yore, let’s go back even further, to the early fire insurance industry. The only sensible way to underwrite a building was for the underwriter to personally visit each prospective property to assess the risk. This approach mostly did the job but was woefully inefficient. 

Daniel Alfred Sanborn began producing and selling fire insurance maps to underwriting companies in the late 19th century. Sanborn Maps provided troves of underwriting details like construction type, fire walls, premises boundaries, natural features, occupancy and proximity to hydrants, fire departments and other structures and hazards. It became possible to underwrite properties without leaving the office. (Don’t listen to the naysayers; insurance has long been a hub of innovation.)

As we progress into the Y2K fears (let’s resist the urge to reminisce about those doomsday prophecies and their accompanying policy exclusion forms), seemingly overnight every insurance professional had a computer. Electronic applications could be completed and emailed quickly to multiple underwriters with a few clicks, and the carriers could expand their distribution more easily.

As insurance companies grew, they fostered heightened competition and developed products that were increasingly customizable to individual risks. 

See also: Underwriting in the Digital Age

One problem persists: The underwriting primarily happens prior to the policy term and is largely based on the historical performance of the insured and other risks of similar makeup. This dynamic is particularly prevalent in small commercial insurance. Small businesses are complex enough to warrant assistance from a trusted adviser to help them get the right coverage at a fair price but small enough in revenue and loss volatility for the carrier to not afford much expense for continuing maintenance. Often, the carrier even waives their right to audit the policy’s exposures.

As we explored in Part 1 of this series, the set-it-and-forget-it strategy for small commercial underwriting has been the standard. Grow the book, ignore it, watch the loss ratio creep, re-underwrite the book. Rinse and repeat. Consumers are at the carriers’ mercy, only hoping they aren’t tabbed to be part of the portfolio getting chopped.

Now -- a case study

Through continuous underwriting practices, we today possess access to data to be smarter underwriters. Through automation, we can scale to benefit more consumers.

To demonstrate how these strategies can be implemented, let’s examine the insurance cycle for Billy Bob’s Bistro (a play on the authors’ names, if you’re paying attention). 

After Behemoth Insurance Company was too aggressive and had to exit the restaurant insurance space, Billy Bob’s Bistro’s insurance agent must find replacement coverage at renewal. Through only a couple simple inputs, the new carrier can instantly learn more about Billy Bob’s than ever before.

As with souped-up Sanborn Maps, the underwriting community consumes assessor data and geospatial imagery to determine construction type and condition, proximity to hazards, fire protection and other underwriting characteristics like replacement cost estimation. We have advanced insights to help us understand probabilities of both natural (flood, hail …) and manmade disasters (crime, nearby underground storage tanks and their last reported leak …). 

Much of this could be considered traditional underwriting data, just accessed instantaneously. But in that same instant we’re now also harnessing additional data. Through Billy Bob’s Bistro’s public profiles on search engines and social media, we can assess things differently than ever before. Reviews alert us of food poisoning concerns, online ratings can show when customer sentiment is decreasing to indicate potential heightened moral hazard, photos can identify the children’s playset that is not properly maintained, the menu’s cuisine points to the equipment maintenance requirements based on the types of cooking performed. 

See also: The Promise of Continuous Underwriting

These insights coupled with a continuous underwriting process are advancing small commercial underwriting into a more sustainably profitable space. With such robust data available, there’s no need to set-it-and-forget-it any more. Now, sophisticated software can monitor changes to individual risks and portfolio trends without increasing overhead.

When Billy Bob’s changes their closing time to 2am and begins promoting live bands, the underwriter knows. When they expand from paninis to deep fried gator bites, the underwriter knows. Because modern technology can make these observations for us, we gain efficiency through monitoring processes.

One source of frustration for Billy Bob was always his term-ending audit. Now, he can register for a pay-as-you-go (PAYG) policy that integrates directly with his point of sale (POS) system. The U.S. restaurant POS market is projected to exceed $10 billion by 2030, thanks in part to the proliferation of food delivery apps integrating into these systems. Because the underwriter now receives Billy Bob’s detailed sales reports in real time, the need for an audit is eliminated, saving Billy Bob time so he can focus on his own customer service. Additionally, because his business has seasonal shifts, the PAYG policy automatically adjusts his premium to reflect his true exposures, a cash flow benefit.

To capitalize on their more popular recipes, Billy Bob’s Bistro buys a van for catering. They add a usage-based auto insurance (UBI) policy to cover the auto exposure while tracking the driving behaviors through a telematics app. Because the trips are infrequent and within a small radius, the UBI policy automatically lowers the premium - but this is more than offset because of Billy Bob’s lead foot.

Billy Bob secured replacement coverage quickly that automatically lowers his premium during slower months and doesn’t force him into a time-consuming audit. His agent has substantially reduced the amount of information he has to provide during the application and renewal processes, allowing the agent to focus on additional revenue generation activity. The insurance company’s PAYG and UBI policies provide an updated risk profile that more closely reflects the desired appetite, removes the risk of premium leakage and guards against insurance fraud with independently verified exposure data.

Most excitingly, these practical examples are all available today with concepts applicable in all industries, including service, retail, hospitality, contracting, healthcare, manufacturing and transportation. With the rapid development of AI, the possibilities for additional insights and predictive models will continue to increase the application effectiveness of continuous underwriting.

Because we are able to process more data more quickly and precisely, we are less reliant on human observation to make every decision. Now those underwriter’s eyes can evaluate a risk with far more granularity. Done properly, this can provide benefits to stakeholders on all sides of the insurance transaction.

Continuous underwriting is here to stay and already proving its worth. Capabilities will only continue to expand as new data sources are brought into the underwriting workbench. Done properly, continuous underwriting can provide benefits to stakeholders on all sides of the insurance transaction and advance the practical evolution of underwriting and insurance, even if it’s hard to hang on the wall next to your firemark Sanborn Map and Y2K exclusion. 

The insurance industry is facing a talent shortage several years in the making at a time when business expenses and policyholder expectations are higher than ever. Many insurance companies are burdened with inefficient billing and premium collections technologies that complicate payments.

Insurers must utilize digital transformation to overcome the workforce shortage, transform the policyholder experience, and improve organizational efficiencies allowing staff to focus on other critical tasks.

Download The Digital Bridge: Closing the Insurance Talent Gap and learn how to:

• Leverage technology to bridge the industry workforce gap
• Enhance the policyholder experience by automating the most frequent touchpoint
• Ease staff workload, reduce costs, and retain policyholders

Sponsored by ITL Partner: InvoiceCloud

Robert Marshall is the founder and CEO of Whisker Labs. Whisker Labs, a spinout of Earth Networks, delivers next-generation home energy intelligence technology to realize the full potential of the connected home.  In 1992, Marshall co-founded AWS Convergence Technologies, the company that would become Earth Networks, by pioneering the networking of weather sensors and cameras using the internet. By developing groundbreaking technology to find "signals" — valuable, meaningful intelligence — in big-data "noise," Marshall improves people's lives and protects their livelihoods. He has appeared on CNN, BBC World News and ABC Nightly News and has been quoted in major news outlets that include the New York Times, the Washington Post, Nature and Scientific American.

Insurance Thought Leadership:

How is Whisker Labs technology making homes and communities safer? 

Bob Marshall:

Whisker Labs' breakthrough technology, Ting, protects homes and communities from fires by detecting arcing, the precursor to most electrical fires. A single, plug-in sensor monitors an entire home and is proven to prevent four out of five electrical fires. To date, Ting protects nearly 450,000 homes and has saved over 6,500 families from electrical hazards that could have led to devastating fires. Outside the home, Ting is advancing how grid resilience is measured and monitored as it detects electrical faults along the utility grid, helping protect communities from catastrophes such as wildfires.  

Insurance Thought Leadership:

Can you discuss some of the challenges associated with integrating IoT technology into insurance and risk management? 

Bob Marshall:

Successful integration of an IoT program relies on overcoming several fundamental challenges to fuel program growth and a robust ROI: 

1. The program must be strategic and have executive buy-in. Support for the IoT technology must start at the top. From there, all stakeholders should see the technology as a strategic investment for their customers and for the long-term benefits it will provide. I should add that, while early forays into IoT largely fell flat, they still yielded lessons on what doesn’t work. In any case, these past projects should not serve as reasons for hesitation. It is a different time, and IoT is here to stay.  
2. The solution must be compelling. Homeowners must be inherently interested in the value proposition the IoT technology offers. In the case of Ting, for example, we find that people rightly have an innate fear of fires. The offer of electrical fire prevention naturally resonates throughout the homeowner community. This drives successful enrollment rates. 
3. There must be a robust, yet turnkey marketing program. To further fuel interest, marketing strategies must be implemented to spread awareness and make signing up easy. An IoT business partner should make it simple for insurers to unlock required marketing channels such as landing pages, email, direct mail and agents engaging with the program. Clear and consistent messaging across all touchpoints helps earn homeowner trust. 
4. The technology has to be super simple for customers. Once successful enrollment is established, ROI hinges on a high activation rate. The technology must be easy to understand, low (no) maintenance and highly compatible in all homes. Ting, for example, is a DIY install that takes just two minutes, which helps drive activation rates over 85%. It turns out our focus early on to wrap the complexity of the technology into such a simple customer experience was even more critical than we first thought – the results validate our approach. 
5. Operations and data integration must be seamless. Behind the scenes, data flows via API integrations must be in place to ensure IoT providers and insurers can seamlessly connect and communicate through all stages, from fulfillment to claims avoidance, and ensure a stellar customer experience at every step. 
6. Finally, ROI must be demonstrated. This first means that the technology has to do really well at its fundamental function. Equally important, IoT partners and insurers must be fully aligned on success metrics, and ultimately the program must demonstrate an ROI that supports long-term success. This, in turn, requires proven technology and a willingness on the part of the insurer to expand rapidly – and the ability of the IoT partner to support that scale – to demonstrate the true impact the technology can deliver.   

Insurance Thought Leadership:

What are the pros and cons, from your perspective as an IoT service provider, of the new software standard Matter? Will it accelerate consumer adoption of devices like Ting? 

Bob Marshall:

We welcome any industry-wide effort that has the homeowner’s best interests in mind. However, all signals point to a slow start for the standard. It needs more time, and it will be a while until other device classes are incorporated and adoption becomes mainstream. At that point, it could help accelerate the adoption of devices that the consumer otherwise would have avoided from the inconvenience or complexity of splitting command and control across more than one hub/interface/app. With Ting, however, fire avoidance is such a compelling value proposition – and there are no command-and-control dependencies – so we don’t anticipate Matter having a material impact on Ting's adoption. 

Insurance Thought Leadership:

Whisker Labs has been able to use data to show utilities the power of analyzing an electrical grid for potential issues. It’s a good example of commercial risk management applications for your device. Can you describe how that use case is going? 

Bob Marshall:

We're finding the use case is even more compelling than we first thought.  Frankly, it is alarming that millions of homes and businesses are being supplied with dangerous levels of power from an aging, deteriorating and often-over-tasked electric system. This problem is only amplified by climate change and increases in electrical demand and grid complexity. The Ting sensor network is the largest network to ever measure and monitor the electrical grid, revealing first-of-its-kind data on its health and state of resilience. The beauty of that for all stakeholders is that the network is already deployed. 

Ting empowers utility companies to think about prediction and prevention; in the same way, the insurance sector has adopted this model. Currently, Ting data is being used to detect and repair home fire risks that originate from faulty utility equipment. But the potential impact is much greater. Data from the Ting network can help to prevent catastrophic events like wildfires and grid meltdowns, as it detects faults along the entire grid and can help direct investment to the areas where it is needed most. 

Insurance Thought Leadership:

Are there other risk management use cases for Ting that you see in the next five years? 

Bob Marshall:

We are laser-focused on helping to protect customers by preventing fires in homes and throughout communities. Ting technology does, however, present several opportunities beyond fire prevention. For example, Ting is already capable of temperature-sensing, which has resulted in cases of frozen pipe avoidance. Similarly, Ting prevents water losses as it detects electrical hazards with sump pumps, water heaters and other devices and appliances that can lead to leaks.  

In the future, Ting will have the capability to predict problematic equipment failures with HVAC systems and other key appliances. The technology can also be expanded beyond the home to commercial applications for insurance partners. Ting’s ability to monitor and spot power quality problems holds great promise for commercial equipment resiliency. Perhaps most compelling is the opportunity to grow and integrate the data sets Ting is collecting with existing insurance risk models to help sharpen and expand how our partners think about and evaluate power and fire risk.    

Insurance Thought Leadership:

Insurers need data to evaluate what reductions in premiums could be offered to those with smart home devices. Has that been a challenge to generate, or is there an easy case to make for Ting? 

Bob Marshall:

Our model with Ting is to provide the technology to homeowners for free. Economic ROI for our insurance partners is clear, and marketing benefits are off the charts, with a constant stream of incredible testimonials from customers who are grateful that the insurance partner, via Ting, has helped to protect their home and family. The benefit of risk reduction thrills customers today even without lowering premiums. It is still a little early for insurers to inform reductions, but with Ting at scale, they will be able to offer more tailored premiums that will likely result in a reduction due to Ting protecting the home and to substantially lowered claim costs. 

Majesco’s latest research report reveals the top 10 trends in 2024 that are shaping the insurance industry and why insurers need to move beyond legacy and adopt new technology that can pave a way for a brighter future.  

Read Now

Sponsored by ITL Partner: Majesco

Vehicle makers have long complained they don't generate the revenues and profits they deserve. They design, build, market and sell the world's cars and trucks, creating a massive market -- but a huge percentage of the value flows downstream to others.

Car makers may capture revenue from financing and perhaps warranties, but others service and repair the vehicles, provide the gasoline, develop the navigation apps and, yes, sell the insurance. The auto makers would dearly love to capture more of that downstream revenue and seem to have especially aggressive plans on insurance.

But a recent article on Tesla's problems developing an insurance business, on top of Goldman Sachs' face plant in consumer banking, suggests that the auto behemoths may not be as threatening as insurers once thought. 

The big original equipment manufacturers (OEMs) have premised their push into insurance on data. At a time when consumers are increasingly open to having premiums priced based on how much, when and how they drive, insurers need to be able to collect that data -- and the OEMs already have access to it.

Because of all the sensors and communications capabilities they build into their vehicles, OEMs don't have to install a telematics device, monitor a driver's behavior for a month or two and wait to generate an appropriate premium for a new customer. The OEM already knows what that car has been up to in prior months and can generate a personalized premium from the get-go. 

But a recent article in Reuters about Tesla's move into insurance shows the difficulties that come when a manufacturing company that touches consumers only indirectly, through dealers or repair shops, tries to set up a direct-to-consumer service business. A consumer business comes with, well, consumers, and those consumers demand service. 

Reuters reports:

"Complaints about Tesla Insurance are drawing scrutiny from state regulators and the plaintiffs’ bar. The Ohio Department of Insurance at least twice this year determined that Tesla had violated the state’s insurance regulations in handling claims, including for a lack of timely communications with a policyholder.... Phil Fioresi Sr., a stonecutter in South San Francisco, California, told Reuters it took about 15 calls to reach someone at Tesla Insurance after his daughter’s car was struck by one of its policyholders in September....

"The insurer wouldn’t divulge the current number of claims adjusters. But the dozen or so adjusters who started handling California claims in late 2021 were quickly so swamped that resolving cases took weeks or months, the people familiar with the operations said. At the time, Tesla insured more than 50,000 vehicles in the state, according to California Department of Insurance records....

"Working out of a Tesla office in Draper, Utah, the initial adjusters sometimes had to take on hundreds of claims each, far more than at other insurers, according to the sources with knowledge of Tesla Insurance’s operations. Unlike competitors that often have separate call centers to take claim reports, Tesla’s adjusters had to answer the phones themselves while also handling claims."

Tesla is also facing a class action that alleges it overcharges insurance customers because its data-gathering is faulty and unfairly generates reports of dangerous driving, as this article in Forbes explains. 

Tesla's CEO, Elon Musk, has been known to fly by the seat of his pants at the many companies he runs, and we will surely see a more disciplined approach from GM, which has announced big plans for insurance, and Ford, which is rumored to have a major initiative in the works. 

But Goldman Sachs still offers a cautionary tale. It has the same sort of strong brand that GM and Ford do and is known for relentless management. It even did a deal with Apple, maybe the strongest brand of them all these days, to offer a credit card. Yet Goldman Sachs has announced it is leaving consumer banking and has taken billions of dollars in losses -- losses totaled $1.2 billion as of August 2022, according to Fintech Nexus, and have continued, to the point that Apple is seeking to end its partnership with Goldman Sachs. 

The problem wasn't lack of customers. Goldman amassed more than $100 billion in consumer assets, and more than 6 million signed up for the credit card. The problem was that, while investment banking and consumer banking are both financial businesses and have similarities when viewed from 10,000 feet, they are very different when viewed at street level, where those consumers live.

Fintech Nexus reports, "One of the main areas of focus of [an investigation by the Consumer Financial Protection Bureau] had been Goldman’s handling of credit card disputes, the level of which the company had been reportedly unprepared" for. 

Goldman Sachs also misunderstood a key issue about consumer banking -- it thought physical branches would fade in importance faster than they have, creating an opening for a generally virtual presence by Goldman. In addition, the firm found that "robo-advisers" weren't nearly as effective in a mass market as it had hoped. 

GM and Ford aren't Goldman, and insurance isn't consumer finance, but Chunka Mui and I found in our years of research into corporate failures for "Billion Dollar Lessons" that planned moves into adjacent markets are one of the seven strategies most likely to lead to disaster. The reason: exactly the sorts of issues that Goldman faced and that GM and Ford may well find. Markets that seem adjacent at a high level can have unforeseen and crippling complications.

Already, there is reason to doubt whether the OEMs will have as big a data advantage as they seem to think they do.

There is a battle shaping up about who owns that data. Insurers act as though they do, whether individually or shared through Allstate's Arity unit. Car makers act as though they do. But I think consumers own the data, and regulators are increasingly siding with me.

The regulators, especially in Europe, are reining in the indiscriminate collection of consumer data by Google, Meta and others and giving control to consumers. And it's hard to imagine that limits set for Big Tech won't filter their way into other parts of the global economy, including car insurance. 

If GM has to ask me for the use of my data rather than just collecting it through OnStar, the equation changes.

Tesla, GM, Ford and perhaps others will, of course, face firmly embedded competition, too. As Barron's reports, "While the opportunity is big, making it a reality won’t be easy. State Farm, Allstate (ALL), Progressive (PGR), and Geico, which is owned by Warren Buffett’s Berkshire Hathaway (BRK.A), control roughly 50% of the U.S. market and won’t cede share without a fight."

Barron's adds that "the data coming from the cars may not be as valuable as the auto companies think. For the most part, auto insurance... shouldn’t be all that complicated. Calculating the number of cars that will get into an accident and what it costs to fix them is relatively easy.... The data coming from cars, while helpful, may not be all that necessary.

“'Seventy percent to 80% of drivers are what are called clean, meaning they just haven’t really had any accidents in five years,' [a research analyst] says. 'You don’t need user [data].'”

Cheers,

Paul

One year ago, ChatGPT was officially introduced and sent the business world, pundits and laypersons alike into a frenzy.

So many aspects of work, life and business are projected to be at risk or, at a minimum, redefined. Conversations are nonstop about the incredible potential vs. man’s possible existential demise, as AI is now considered much closer to parity with human thinking than previously thought possible. What happens once AI catches up to and then surpasses human thinking is hard to fully imagine.

At a high level, AI encompasses; machine learning (ML), deep learning, generative AI, large language models (LLM) and the current favorite: generative pre-trained transformer (GPT). For purposes of this article, we will not attempt to explain these further.

AI in Insurance

Insurance is no exception, as technology providers are sprouting up, or, rather more commonly, solution providers are highlighting their existing AI capabilities. The AI vendor community to the P&C industry is rapidly expanding and may generally be grouped by use case; hyper-automation, insights, image and language.  

Some have been at the AI game for longer and may even fortuitously have “AI” in their brand identity. Others have been quick to point out that their work has been surrounded by AI for years, touting both expertise and subject knowledge. Even insurers themselves are experimenting, setting up AI safe zones, establishing so-called red and blue ocean strategies or simply creating AI best practices as a foundational starting point. It is doubtful that any insurance carrier boards of directors or C-suites have not set some AI work in motion. Insurance regulators are attempting to get ahead of things with proposed AI ethical standards but in reality, are in catch-up mode.

Famously, “technology in search of a problem” rarely is a successful approach, and, so far, generative AI for insurance feels more like a technology seeking problems to solve.

There are few AI experts with extensive knowledge and lots of business people with just basic AI knowledge, so components like machine learning, computer vision, large language models, and generative AI can easily get mashed up together. Fortunately, the experts are openly explaining the differences and providing the details on how this all works, and the webinars and conference events are helping the cause. In the meantime, such AI expertise shortages only complicate insurers’ vision for clear use cases, business purposes and ultimately ROI.

While insurance automation is not new, the prospects of applying AI to partially or completely replace humans quickly gained attention. More recently, these views have been tempered with the idea that AI will be better applied as a “co-pilot” for most insurance functions within underwriting, pricing, claims, sales and possibly others. The industry is in the crawl stage of crawl, walk run.

See also: 5 Ways Generative AI Will Transform Claims

AI for Claims

Conventional wisdom is that AI lacks human emotion and empathy. Claims might demand the most human emotion in insurance, so the AI use-cases talked about today tend to call for AI tools aiding claim adjusters rather than doing the whole job. However, all the discussion is still early and focused on the short term. ROI still dominates decision making and, given the highly competitive P&C insurance market, fraught with financial pressures, the balance between deployment of tools and automation of jobs will be put to new and more rigorous tests.  

Underwriting and claims emerge as the top use areas, which makes sense. Large amounts of data are used to assess and price risk, and, similarly, claims is all about gathering information and making decisions. Both functions are people-based and are already pursuing automation agendas like low-touch and straight-through-processing.

Within the claim space, much of the generative AI talk is heavily weighted to reviewing and summarizing records, such as medical billing or a demand package. The overarching wisdom is that claim handling is record- and paper-intensive. A common misperception is that all claims are alike. Claims insiders say, "A claim is a claim,” but that is misleading when applied broadly. High-frequency/low-severity claims differ greatly from the most complex claims that happen infrequently. Some 70% of auto claims, for example, have minor to modest damage and no or only minor injuries, with few documents to summarize.

Where Can/Should AI Be Applied Today?

The good news is that AI in claims is already successfully being embraced. Computer vision for total loss prediction and photo estimating is far-reaching. AI fraud models are helping carriers scan for anomalies for investigation. However, even within the best AI claims examples, there is a long way to go to reaching meaningful ROI. Trepidation around fairness, legal and regulatory pressures and data security when training models are valid concerns. Even so, there is room for more creative use-case thinking, and the following is a small sample of possibilities. 

This is a sort of AI use-case wish list, free from prioritization and not exhaustive but widely appealing because these are among the most repetitive and demanding claim functions:

• Claim intake for assignment accuracy, reducing or eliminating reassignments
• Claim triage
• Fraud detection, especially organized fraud
• Categorization and severity
• Coverage guidance
• Comparative negligence determination - which party(s) are at fault and to what degree?
• Correspondence generation
• Injury and damage evaluation 
• Settlement recommendations
• Notes analysis and summarization
• Case reserve and formula reserve setting or reserve portfolio management
• Regulatory compliance in real time
• Regulatory reporting: Summarize, validate, review and report
• Pending claim management, prediction and prioritization
• Business interruption claim analysis
• Quality assurance review/auditing
• File summarization for management review, file and settlement authority
• Productivity management measurement

See also: Insurers Boosting Their Use of AI

There certainly are risks to balance when it comes to the degree of co-piloting or replacing people. While there is excitement for automated and AI powered claim customer services, there is a natural dependence on chatbot acceptance to overcome, not to mention room for the connected claim ecosystem to become truly connected and coordinated to realize gains.

Moving ahead, carriers will need to apply additional filters when advancing use cases. Generally speaking, insurtech, including AI, falls into efficiency gain/expense reduction emphasis by automating process and reducing full-time employees (FTE). The elephant in the room today is insurer profitability from soaring indemnity costs in which there can be far greater influence from loss ratio improvement compared with loss adjustment expenses (LAE). Yet the P&C industry has overemphasized LAE reduction because of the simplicity in measuring operating costs.

Insurers will continue to buy vs. build AI through integration partners as a way forward. Solution providers will need to move closer to unravel high-value use cases. The open challenge to insurers and AI solution providers is coming together to develop meaningful business cases, including loss avoidance, mitigation and payout accuracy beyond efficiency gain.

According to a recent Slack survey, a staggering 43% of desk workers report feeling burned out. This problem is especially pronounced in the insurance industry, where the burnout rate stands at 39%, surpassing the national average of 35% and making it one of the top five sectors reporting the highest levels of burnout in 2022. 

While the effects of burnout are felt by every member of an insurance organization, insurance agents are particularly vulnerable. The role of an agent involves selling policies and catering to the needs of existing policyholders, requiring them to interact with many individuals and organizations on a daily basis. If an agent becomes overwhelmed, their ability to fulfill these critical responsibilities may be compromised, leading to negative consequences for both policyholders and insurance carriers, including reduced customer retention rates, decreased revenue streams, and lower satisfaction scores, among other adverse outcomes. 

Today’s insurance talent gap may be pouring gas on the already-blazing burnout fire, but there are ways organizations can bridge the workforce gap and better enable their agents to enjoy their work and be more productive. 

Read More Here

Sponsored by ITL Partner: InvoiceCloud

While the decade is all about tech transforming insurance, the year 2023 has definitely been all about Generative AI. As agencies connect individuals and businesses to the right insurance products - they will need all the help they can to provide superior experience to customers, forge stronger relationships with clients and increase sales. And the right toolkit is crucial to making this happen.

Here are four areas agencies should consider ‘technovating’ to boost their business in the coming years.

1. Empowering Your Agent: Is your agent toolkit really helping them sell? The topmost priority for any seller tool is a mobile-first design. An agent is out building relationships, so storing all meeting information and returning to feed this into a CRM is a killjoy. It stymies productivity and lowers the enthusiasm of agents who enjoy the process of selling more than anything else. Think of tools that provide your agents a seamless mobile experience, with a single interface for all their needs - tracking client locations ahead of a meeting, searching for the right collateral to share at the next meeting, catching up with the manager for a quick chat before a meeting, capturing meeting notes, setting up reminders for follow-up. All done easily, quickly, and effortlessly on the mobile, via a single screen. 

Your agent will be ecstatic! 

2. Nailing the Renewals Business: Now we all know that revenue from renewals are much higher than that from new business. So ensure that the tech you are using actually helps you strengthen your renewal process by,
   1. Auto-calendaring and tracking renewal dates of all products across your database
   2. Recommending the right time to follow-up on renewals based on previous history
   3. Nudging and reminding agents to engage with customers optimally

Simple fixes like this can ensure ~3x increase in revenues.

3. Focusing on Learning and Development: With the global insurance industry trying to navigate competition, regulatory challenges and the need to innovate, it is important to weave in a culture of continuous learning within the organization to help agents and relationship managers stay ahead of their game. 

Is there a way to integrate learning into the same tool that your agents are using? Leveraging AI and ML can actually help build a strong database of winning agent behaviors and best practices which is a dynamic way of coaching agents while they are on the go. Nuggets of insightful advice and mentoring from tenured agents, managers, star performers can harbor a thriving learning culture across the agency. 

4. Optimizing your Lead Management: A good way to rein in business is to ensure airtight lead management. Typically 5% of your leads convert. What are you doing about the remaining 95%? This is typically lost out due to various reasons. But here are the reasons you can actually fix:
   1. Leads that are not activated at the right time
   2. Leads that slip due to poor nurturing
   3. Leads that go sideways due to wrong allocation

AI and ML based technology are built to address these challenges. From pairing the right lead to the right agent based on several parameters, to nudging the agent to call them within the recommended time period, to following a strong lead nurture process backed by data insights, organizations have reported a 25% increase in conversions.

As we step into 2024 and chart our business goals, investing in the right tech can help in two ways: streamline and organize your internal processes and journeys, engage your agents with intuitive, easy-to-use technology,  and provide a roadmap to better revenues. 

Sponsored by ITL Partner: Vymo

The insurance industry is no stranger to data. Insurers need all the customer data they can get — structured data on a policy form, past behavior of policy renewals or terminations, and publicly available government data such as criminal records, bankruptcies, and foreclosures.

While insurers have access to mountains of data — everything from disaster models to historic perils, real-time weather feeds, and current policy information — they don’t have an effective way to integrate all these disparate data sources into one comprehensive view. Until we get there, we won’t have the deep insights we need to prevent risk or develop new products.

The faster we move as an industry to digitize that data, automate underwriting, and integrate third-party data, the more quickly we can create new underwriting and portfolio risk management techniques.

However, the silver lining is that some insurers are already integrating big data and analytics into their operations, from creating early-warning systems to gathering insights that prevent incidents to simplifying and accelerating claims processing.

Here’s how you can reduce friction throughout the customer journey — from requests for coverage to claims — while mitigating risk and cutting costs.

Risk-Based Pricing Innovation and Transformation

Even without face-to-face customer interactions, you can build accurate customer profiles by aggregating data from various sources — and scoring this based on demographics, social media, and public-facing data — to create more personalized products and risk-based pricing.

We’re seeing this play out in the auto insurance industry, which traditionally is highly competitive and price sensitive. Customers are known to switch insurers on price, which impacts growth and profitability.

However, with the help of customer data — structured and unstructured (reviews and ratings, blogs, social media, and auto forums) — you can better understand the customer, make correlations, categorize the risk status of different models, and set premiums accordingly.

Armed with these insights, you can:

• Accurately assess each unit’s risk within an insured pool by micro-segmenting risk pools.
• Price dynamically based on changing risk profiles — all enabled by granular behavioral data and advanced modeling.
• Offer competitive and differential pricing while improving premiums and demanding higher premiums for riskier products.

For example, in Spain, Zurich Insurance launched Klinc, the innovative, on-demand insurance solution that lets users of web and mobile applications buy cover whenever (and wherever) they need it.

In the future, we’ll see more cross-industry platforms and seamless insurance portability that integrate risks and exposures and covers for an insured under a universal insurance policy.

Point-Based Risk Scoring

The greatest challenge for commercial insurers is collecting data — policy in-force and industry-wide peril data — from disparate sources. Risk analysts constantly need context to make critical business decisions.

Enter location intelligence.

Location intelligence is the lens through which analysts, underwriters, and actuaries can view complex data, uncover opportunities, and identify hidden risks. Because of the spatial connection between policies and perils, location intelligence helps analysts extract and share actionable insights throughout their organizations.

By leveraging location, carriers can integrate data from policy databases, spreadsheets, and third-party vendors to get a better view of their portfolios.

This geographic view of the data provides:

• An intuitive platform for identifying clusters of policies in effect
• Concentrations of insured value
• Metrics on previous claims

With deeper contextual insights into potential perils, analysts can determine the exposure of an individual policy to perils on a hyperlocal level.

Location intelligence has transformed actuarial science paving the way for point-based risk scoring, the underwriting method that examines a policy’s relative exposure to the perils that may occur.

Point-based risk scoring is a drastic shift from the traditional method, where rates are determined based on whether a policy existed within a risk zone.

There have been instances where flooding and fires have occurred outside traditional risk zones, such as Hurricane Harvey and the Fort McMurray wildfire. But it’s still a more effective risk-scoring method that accounts for perils that could occur but have yet to happen.

Take, for example, a car owner who lives on a corner lot flanked by a busy road on one side and a quiet street on the other. The risk associated with this location would come down to which direction the driveway faces and insurers would then adjust for it.

Trusted data with location context allows insurers to price accordingly, mitigate risk exposure, and improve reinsurance rates.

Intelligent Risk Assessment

Commercial risk assessment is data intensive. A fixed data set is traditionally used to perform a risk assessment, which can be limiting. But when you integrate real-time data from external sources — more granular and up-to-date than you would typically find — you get the complete picture including:

• The likelihood of everyday perils at the property location
• The existence of high-risk items nearby
• The density of existing policies near the property

It helps you better assess the property submission before the policy is accepted.

Individual risk assessment results in special pricing of the premium cover required to cover the risks inherent to the particular property.

For insurers, it means identifying areas of reduced exposure, better risk management, and lower costs. For policyholders, it means better prices, as insurers will have a way to identify regions where they can offer more attractive terms and reward good customers with greater pricing accuracy.

By integrating critical risk-assessment data with historical results, you can quickly identify anomalies in likelihood and impact ratings and assess the geography and product or business lines driving such changes.

To learn more about how you can achieve success by leveraging data analytics and AI, get in touch with us. Read our e-book — Under the Hood: Unlocking the Hidden Value in Insurance Data for all the ways you can reach your digital transformation goals with stronger data.

Murray Izenwasser, Senior Vice President, Digital Strategy

At OZ, Murray plays a pivotal role in understanding our clients’ businesses and then determining the best strategies and customer experiences to drive their business forward using real-world digital, marketing, and technology tools. Prior to OZ, Murray held senior positions at some of the world’s largest digital agencies, including Razorfish and Sapient, and co-founded and ran a successful digital engagement and technology agency for 7 years.

Sponsored by ITL Partner: OZ Digital Consulting