Trends and Advances in Accelerated Underwriting and Current Use of Artificial Intelligence Solutions in the Life Industry
Life underwriting has changed significantly in recent years with the use of Accelerated Underwriting (AU) and other tools, focusing on making life insurance faster and less invasive for applicants while maintaining reliable mortality outcomes.
In addition, the life industry is in the early stages of using Artificial Intelligence (AI) to further drive changes and enhancements to the underwriting process.
To gain insight into current AU trends and the industry's use of AI and other tools, we asked three leading underwriting experts to share their perspectives on what they see happening today and what we may see in the future.
Our panel of experts:
Katy Herzog, Certified FALU, CLU®: AVP Underwriting Pricing and Client Support, Munich Re Life US
As AVP, Underwriting Pricing and Client Support at Munich Re, Katy, along with her team, works directly with carriers to develop, assess, and optimize their underwriting programs. Katy collaborates with various disciplines to deliver process enhancements and innovative underwriting tools while supporting risk management expectations.
Joel Larson, CLU®, FALU, FLMI, MBA: VP & Chief Underwriter, Life Underwriting, Prudential Financial
Joel's responsibilities include setting risk selection standards, which involves defining underwriting rules, requirements and policies; overseeing transformation initiatives; evaluating and implementing new technologies, processes, and data; and leading teams of large case underwriters, technical underwriters, and medical directors.
Kristin Ringland, FLMI, FALU: Senior VP and Chief Underwriting Officer, SCOR Global Life Americas
Kristin leads the underwriting department for SCOR Global Life in the US. She has been with SCOR for 19 years, with 30 years of underwriting experience. She is past President of the Kansas City Risk Selectors, past member of the AHOU planning committee and ALU education committee, and she is a member of the ACLI Risk Classification committee.Here are the questions we posed and their responses.
Since its introduction, Accelerated Underwriting (AU) has become widespread in the life insurance industry. What major trends have emerged regarding face amounts, age limits, and acceptance and offer rates?
Joel:
Overall, the general trend with AU has seen a shift to a more streamlined and data-driven underwriting process. From an industry perspective, we've seen the trend of increasing face amounts and even some expansion of ages considered, from age 50 up to age 65 or even higher. We continue to refine our criteria as more tools and sophisticated data analytics provide additional clarity to the risk assessment. Ultimately, our goal is to offer the most competitive and efficient underwriting process while at the same time ensuring accurate risk assessment. Numbers show that if we're able to make an offer more quickly, our acceptance rates increase.
Kristin:
AU programs were developed to improve the customer experience, provide a faster issue process and reduce the cost of medical tests. The programs targeted rates close to fully underwritten rates, with the expectation of similar mortality results. To accomplish these goals companies initially developed programs that were intentionally more conservative with the idea of bringing an AU program to market but keeping a close eye on potential impacts to mortality. Close monitoring of program results was required to ensure there wasn't significant mortality slippage. Program monitoring has been and remains vital to identifying slippage and improving pass through rates.
With better understanding and confidence in their AU programs, paired with improvements in automation and instant data sources, companies were able to offer coverage at higher face amounts and older ages. Specifically, the availability of instant medical data through Medical Claims Data, Lab data and Electronic Health Records (EHRs) provided for program expansion and in some cases allowed for lower substandard rating consideration.
Katy:
In the decade-plus that accelerated underwriting has existed in life insurance, we've seen continuous evolution with varying degrees of pace and impact. I'll start with a definition of AU as it varies by company and may include fluid-less, non-invasive, and/or automated underwriting. At Munich Re, we define "accelerated underwriting" or AU as the waiving of traditional underwriting requirements (e.g. fluids, medical exams) for a subset of applicants that meet favorable risk requirements in an otherwise "fully underwritten" or FUW life insurance process.
In the early days of AU, we saw rapid progression with overall program implementations peaking in 2018. More recently, it appears to be stabilizing. Per our most recent industry AU survey, average maximum eligibility ages are hovering just under 60 years old. However, eligibility limits continue to increase with an average maximum face amount of approximately $2.5 million. It is important to note that expanded eligibility limits are often accompanied by lower acceleration rates and more human underwriter review.
A key trend has been the emergence of digital health data (DHD) tools. We saw high adoption of these tools in direct response to the COVID-19 pandemic because in-person exams and medical records became difficult to obtain and carriers looked for alternative ways to assess risk. Interestingly, that growth has not slowed down with the largest shifts in medical claims data (Dx), electronic health records (EHR), and aggregated laboratory (e.g. clinical lab) results. On the flip side, it appears fewer carriers are leveraging tele-interviews in favor of e-applications. And although credit data continues to be utilized, there's been a shift from routine to more occasional use.
When we look at key metrics, offer and acceptance rates appear modest in the early days, but they have gradually increased as carriers have become comfortable with their results. Based on our 2024 survey results, on average, 70% of applicants within a carriers' age and amount eligibility limits are eligible for AU. Of those eligible, we're seeing an average acceleration rate of 45% with an average offer rate of 39%. For those that received an offer, the average placement rate in 2024 was 75%. There does appear to be a tradeoff between acceleration and placement rates as carriers with lower acceleration rates reported the highest placement rates, and vice versa. In addition, higher face amount bands are often accompanied by lower acceleration rates. Finally, distribution may also be a differentiator as carriers utilizing agents had the highest placement rates versus direct-to-consumer (DTC) carriers who reported the lowest placement rates.
What new underwriting tools, data sources, and practices have facilitated AU? Are these tools applicable and/or have they been applied beyond the typical age/amount limits of AU?
Kristen:
Some of the same tools are being used, but with significant modification and improvements. Automation has been a part of underwriting for over 25 years. Significant upgrades in technology allow a combination of predictive models and rules in decisioning through automated engines. The models and rules now incorporate instant medical data including clinical lab history results, medical records in the form of APS and EHR, and medical claims data. Over the last several months, we have reviewed and consulted with several carriers on the updating of their applications. Many of these updates incorporate multiple behavioral economic principles and techniques which has proven beneficial in improving disclosure and identifying potential misrepresentation.
The same tools and data sources being used for AU are being incorporated into the fully underwritten space. Many companies use a triage process with a rules engine or predictive model to evaluate each underwriting requirement received and provide instant feedback to the underwriting workbench. AI is also significantly impacting both the AU and fully underwritten process, providing speed and efficiency, as well as consistency in decisioning.
Katy:
I mentioned the growth we are seeing with DHD earlier and I do believe this has enabled AU expansion. Prescription data has become a stable AU tool and has fully saturated the market. But we've seen the biggest shift (since 2018) in medical claims data (Dx) usage, followed by EHRs and then clinical labs. This trend can be attributed to various factors, including the increasing availability and quality of data as well as the environment created by the pandemic.
For EHRs specifically, carriers can obtain traditional-type medical data in an accelerated fashion. Some EHR providers are delivering approximately 90% of EHRs within two days and up to 50% within one hour. In addition, carriers view EHRs as one of the most protective tools for AU. However, EHRs continue to be used more occasionally rather than routinely in an AU setting. This indicates that challenges remain, including hit rates, data completeness, and ease of use. And the industry continues to address these challenges. We've seen continued work on expanding data access as well as streamlining EHR reviews, whether it's a summary tool for a human being to read, automation tools, such as alitheia's EHR assessments, or both, including Munich Re's EHR Summarizer.
Many of these tools are being used in both AU and FUW settings. In fact, many carriers initially look to EHRs to replace attending physician statements (APSs). However, there have been mixed results. Munich Re completed a series of comprehensive retroactive EHR studies and found that while there are opportunities for utilizing EHRs in lieu of APSs, there are clear benefits to utilizing EHRs in AU, as well as fluid underwriting (when an APS is not required).
Joel:
Outside of now using big data and predictive analytics as a conduit to risk assessment, some very specific data sources have emerged as particularly impactful. More and more companies are using medical claims data as an addition to longer-standing pharmacy database checks. Claims data has helped in getting a more comprehensive view of an applicant's health profile and risk which may help in reducing the need for traditional medical exams or medical records. Underwriters are able to leverage this data to make decisions more quickly by making risk-smart decisions when they can collect enough data with this tool and others.
Electronic Health Records continue to evolve and are more widely accepted by the industry. We can obtain this data instantly or in days vs. waiting for weeks for a traditional APS. I really view this tool as the next big leap to expand AU opportunities as they can provide a more detailed and organized view including diagnoses, treatments and medications which allows an underwriter to make a more accurate risk assessment. Currently EHRs do have some limitations with data completeness, but this continues to evolve.
Both medical claims data and EHRs may be used beyond AU parameters as the claims data can be protective across all ages/face amounts and EHRs can be useful as an APS or exam replacement in applicants with less complicated histories.
One challenge distribution encounters with AU is that different carriers implement their programs in slightly varied ways, leading to confusion in the marketplace. How can distribution differentiate between them? At what point in the sales process should the availability of AU be evaluated, and to what extent can a producer rely on this process without affecting the outcome of the case?
Katy:
That's a great question. It definitely involves a shift as it is more difficult to confidently provide the risk class up front. I believe the availability of AU should be accounted for up front. The biggest key is to set appropriate expectations, and to do this you must be knowledgeable in the various programs. Carriers may have different parameters, including AU eligibility, automation, and instant decision capabilities. Some carriers include substandard risks, others do not. If a carrier has knock out medical conditions, they often will publish or share those in their field guides.
AU definitely generates some new challenges for distribution, but it also enables less invasive, faster, and ultimately more customer-friendly processes. When in doubt, ask an underwriter. We want to help and make the process as frictionless as possible for all.
Joel:
I can certainly empathize with market confusion noting the wide swing in the parameters of these AU programs. Traditionally, carriers used to compete mostly with products and the underwriter offer. In today's world, the overall customer acquisition experience is just as important in many cases, especially those in the lower face amounts and ages where a quick and efficient process is necessary. A more efficient process is necessary for carriers to lower acquisition costs and help maintain competitive pricing. I believe that this is just as important for our distribution partners so they can spend their time on the higher face amount cases. As distribution would educate themselves surrounding competing products, a certain level of knowledge about the varying programs is also essential.
The goal is to make decisions earlier in the process which can only help both the carrier and distributor. As more and more quality data surfaces and matures (i.e. EHRs), faster decisioning will follow.
While AU helps make decisions more quickly in many cases, it's never a guarantee. I'd encourage distributors to show this as an opportunity, but I'd steer clear of promising an AU offer as all data may not be available or known at the time of application. While there may be guarantees of an AU offer with certain products or processes, additional mortality charges are typically baked into the pricing to account for the increased mortality slippage.
Kristin:
As a reinsurer, we have reviewed dozens of AU programs in the last 10 to 12 years. No two programs are alike. They vary from the goals that started the company down the AU path, the types and numbers of data sources used, whether they use automation with rules sets, predictive models or both, whether underwriters are involved in the final approval process and a host of other variables. One of the biggest impacts to the mortality of a program is the market of the company. Knowing the market focus of the carrier and understanding the AU process and guidelines is key. Does the carrier have a pre-qualification process or a different application requirement for AU, do they "knock out" cases from AU with no pivot to fully-underwritten and a host of other information you need to be aware of when determining which AU program will work best for your applicant. If in doubt, ask questions.
What role will Artificial Intelligence (AI) play in the future of AU?
Joel:
AI is poised to play an increasingly significant role in the future of Accelerated Underwriting (AU).
For years, AI has been employed through predictive models in underwriting, enhancing efficiency and accuracy. As for generative AI and large language models, the industry is just beginning to explore their potential. These tools offer the ability to significantly improve the underwriting process by swiftly summarizing and analyzing vast amounts of data, including extensive medical records. Currently, large language models are in the early stages of being used to summarize and interpret these datasets and provide rate class suggestions to underwriters. As these technologies mature, there is potential for more complex cases to be processed through the AU system with minimal or no human intervention.
Kristin:
AI is already impacting underwriting in significant ways, whether internally or through partnerships with reinsurers or vendors. Companies are using AI for everything from informal pre-assessments, application review, claims evaluation, pricing and treaty reviews, underwriting and post-issue auditing. AI is not being used in the underwriting space for decisioning, but to assist in every other aspect of the underwriting journey to the point of decision.
Katy:
Artificial Intelligence (AI) will increasingly play a role in underwriting, including AU. Many carriers and reinsurers are already using AI today, although in a controlled or limited setting. Even today, AI can help us discover and analyze the differences between carriers' AU programs in support of the sales process. For many folks, including distribution and underwriting, I think it's important to note that while AI will create efficiencies and may perform certain functions of the job better than we do today, I do not believe it will completely replace humans. For underwriters, it is a tool that will enhance our abilities to assess cases and improve processes while enabling us to focus on the more challenging and meaningful underwriting aspects. I do think that now is a good time to responsibly experiment with AI and learn about its capabilities, uses and limitations so that you and your companies can benefit from it now and in the future.
Where is the industry in using AI and how is it being used today? What obstacles are faced in its development?
Kristen:
The industry is in the very early days where the use of AI is concerned. All the future uses and impacts of AI for insurance, or any other industry can't be known at this time. It is amazing how fast the insurance industry has adopted AI technology. If you've been part of an AI project or even tried to use AI on your own, the technology is truly groundbreaking.
For the underwriting space, AI can be used for every part of the process from taking the application through the final decision. One of the most common use cases for AI our clients mention is the use of a case summary. Underwriting workbenches are being updated to accept the different data points AI can pull from the case documents, the valuable information that leads to a decision. Using AI technology to review the case documents and summarize the identified medical/non-medical information and linking that summary to an underwriting manual for a suggested case assessment provides greater efficiency by means of improved accuracy, consistency, time savings and insightful risk analysis.
Obstacles faced by the underwriting community with the development of AI technology are similar to the challenges experienced with use of predictive models: potential bias and discrimination, regulatory considerations, transparency and explainability, data privacy, etc. Companies must have a clear AI governance framework outlining processes, standards and security requirements to ensure AI systems and tools are safe and ethical. Establishing these guardrails for the development of AI should precede work in the AI space. Ongoing monitoring and evaluation of any AI tool or resource and the changing regulatory landscape will make AI development more responsible and efficient.
Katy:
I believe our industry isn't too far behind others with its use. I see it being used in front and back office functions; however, for certain functions, such as underwriting, AI is being deployed in specific or limited instances as folks test and learn. Per usual, our industry is highly regulated and so I'd recommend approaching with best practices in mind, much along the same lines that we have for predictive models:
- Be smart and thoughtful – have defined use cases and goals
- Implement a governance framework to include bias testing
- Be cognizant of hallucinations
- If using for risk selection decisioning, must be explainable
- Monitor and update
Joel:
While AI can be an incredibly powerful tool, it's important to ensure it's used appropriately as use cases increase. Big ticket issues like regulatory compliance, bias and fairness, and data privacy all need to be considered with expanded use. In underwriting specifically, there is a need for transparency to avoid the "black box" problem, where AI-driven decisions are difficult to understand or explain. Solving this issue is crucial for meeting regulatory requirements and effectively communicating decisions to sales partners and customers.
Additional challenges include the need for expertise and integrating AI with organizational systems which can be costly, complex and time consuming.
Any closing thoughts?
Katy:
It's an exciting time to be in this industry. We've seen continuous progression and innovation, but I believe we're really on the cusp of a major transformation with AI. AI will change all aspects of our industry including sales, underwriting, and customer engagement. As customary with our industry, there will be considerations including regulatory, data governance, and bias reviews. However, I am confident AI will enhance the ways we work and improve outcomes for not only the companies that we work for (carrier and distribution), but ultimately and most importantly, our customers. I'm excited to be a part of this transformation and hope you are as well. We can help shape what this industry will look like and what we can deliver in the future.
Joel:
As an underwriter, it's understandable to feel some apprehension about the capabilities of AI. However, I choose to see it as an additional tool that can enhance our abilities. I'm enthusiastic about the potential for AU programs and AI to further transform our profession by alleviating some of the routine tasks like condensing voluminous amounts of data or even making more routine decisions. Human expertise remains crucial for oversight, contextual understanding, and bringing creativity to our underwriting offers. It's important that companies maintain distribution relationships, especially with more complex and larger cases. Making an underwriting offer is one thing but being able to explain or "sell" that offer will always be vital. After all, this is still a relationship-driven business.
Kristin:
Personally, I'm excited about what the future of underwriting holds. Technology and innovation in the underwriting space is moving at a fast pace. This is a good time to embrace the changes, paving the way for future success.
Comments