LIMRA Annual 2024 Recap - AI is a Tool, Not an Employee
LIMRA Annual 2024 was full of AI buzz. Will AI replace jobs? Who is leading the pack in the insurance industry? Who is behind and how long can you wait to mitigate risks?
The answer: AI is a tool, not an employee. And the wielders of powerful tools built to solve niche problems typically succeed.
The State of AI in Insurance
According to LIMRA, investment in AI is high and growing, with 100% of carriers experimenting with GenAI solutions and 75% with active or planned pilots and implementations of AI. However, this by no means suggests that AI adoption is mature. In fact, AI is in its infant stages when it comes to carrier adoption, with many challenges highlighted by prominent carriers such as change management, eradicating enterprise silos, data integrity, legacy systems, and more.
A central theme at the LIMRA Annual Conference was that AI is a tool, not an employee. Strategically, insurance executives are aiming to utilize AI tools to drive efficiency gains, better products, and growth within their organizations. History has taught us that innovation resulting in better products and outcomes for consumers leads to market growth over those who do not adopt new technologies. Hence, AI adoption can be seen as a growth opportunity to capitalize on what David Levenson (pictured), CEO of LIMRA and LOMA, exclaims is a massive unmet demand for products like life insurance in the US. Workplace benefits alone grew ~8% from 2022 to 2023 to a whopping $8.3B.
How to Build Powerful AI Tools
One way to think about AI is by comparing it to a scalpel. Like a scalpel, AI is highly precise and powerful when used for a specific purpose. A scalpel can deliver intricate solutions and transformative results, but if you try to use it for a task it wasn’t designed for, you risk doing more harm than good. Similarly, AI is not a one-size-fits-all solution; it works best when applied to well-defined, specific challenges in insurance, like automating claims processing or improving underwriting accuracy.
1. Start with Specific Problems, Not with Flashy Technology
It is no coincidence that the masters and news headliners of GenAI - OpenAI, for one - were simply not present at LIMRA. This is because the problems these large language models are solving are largely generic. However, in the insurance industry, it is important to first focus on the specific problems you are trying to solve for, rather than become lured by flashy technology built for the masses. Ultimately, the goal is to find the AI Tools that help you achieve your business goals and solve your specific problems, and not to pick up the largest tool in the toolbox and try to wield it. These problems include decreasing costs by driving efficiencies. In fact, in workplace benefits there is a 2x to 4x difference in cost between top performing carriers and bottom performing carriers in operations, underwriting, IT, and more.
As was eloquently stated in the workplace breakout session "At the Precipice of Possibilities: Innovating with Purpose" with David Payne, Michael Estep, and Sean O'Donnell: "Technology is enabling innovation, but it does not necessarily drive it." The leaders who wisely use these new technologies to solve their problems and grow their businesses are the ones who drive successful innovation, and that starts with defining your business goals and building the appropriate tools to achieve them.
2. Data Readiness
Data readiness and compliance was a large open discussion at the conference. During the panel "Unlocking the Promise of AI: Navigating Data Readiness Needs," data readiness was quoted as "one of the biggest challenges to prepare insurance companies for AI." This is so critical because AI depends on data quality in order to become a powerful tool for your organization. Due to the general scattering and sometimes inconsistent data across different systems, carriers must strategize how to organize and collect data in ways that can be made accessible to AI models. Many also pointed out that compliance with data privacy is essential when using personal data to train AI models. AI applications in insurance cannot compromise customer trust, and mishandling sensitive data can lead to breaches or regulatory penalties.
One such way to handle data readiness that was discussed at the conference is to gather and carefully label data in a secure vendor platform designed to solve a specific problem. For example, Uniblox's new AI platform enables carriers to unlock new efficiencies by training AI models on existing group census data that is loaded into the platform and validated from day one. In this way, data quality, accessibility, and compliance remains high without having to start from scratch or extract extraneous data from a carrier's data lake, which may be hard to access and may contain undesirable or conflicting data that would run counter to your business goals.
3. Enterprise Architecture
Ultimately, a general agreement amongst carriers and vendors at the conference is the desire for enterprise-grade, insurance-specific architecture as everyone continues to weigh the risks between speed and regulation in the world of AI. Working with vendors who can help you design enterprise-grade AI architecture to achieve your business goals is critical for keeping trust with all of your stakeholders and key to avoiding pitfalls from generic models. Those who design powerful AI tools built to solve niche problems with quality data will be able to harness the true power of AI and will become competitive leaders in the industry. Moreover, cybersecurity measures are critical to protect these systems from external threats, ensuring that AI models and the data they rely on are secure from potential breaches.
How to Use AI Tools Successfully
Once you begin building and piloting new AI tools, using them is the only way to extract the real value from your investment. Step one was echoed often at the conference - change management. One executive wisely pointed out, "what is often understated is how technology that is brought along to aid certain departments does not always directly help the people who first start implementing the change." With this in mind, leadership during these times is critical to keep everyone motivated towards short and long-term business goals that help the entire company move forward and grow.
Additionally, as AI becomes more integrated into business processes, governance cannot be an afterthought. AI governance frameworks should establish clear policies for accountability, transparency, and bias mitigation, ensuring that AI solutions comply with legal requirements and uphold ethical standards. Several carriers at the conference discussed plans to create internal AI governance frameworks to ensure the responsible and compliant use of AI tools within their organizations.
Summary
At LIMRA Annual 2024, AI emerged as the central theme, not as a threat to jobs but as a powerful tool. The future of AI in insurance isn’t about replacing human workers; it’s about solving specific, niche problems that enhance efficiency and drive growth. However, success lies in precision: those who apply AI like a scalpel—carefully, thoughtfully, and with a clear purpose—will be the ones who lead.
Data readiness, enterprise architecture, and AI governance are critical to building reliable AI tools that scale effectively and responsibly. The leaders who wield these tools with precision and focus will define the next era of innovation in insurance.