Takeaways
- The AI conversation isn't about adoption anymore but acceleration.
- AI can enable insurance companies to help customers in new ways.
- Insurers with legacy systems designed for humans to use, not for AI, could have challenges.
Responses have been lightly edited for clarity.
How is AI being used to influence the customer experience?
The most obvious to most people, is when they want to call into an insurance company. What happens after hours? What happens if they don't speak the same language? So, at the very basics, AI enables us to serve customers in new ways we haven't had before. AI-powered chatbots are so much more intelligent than the traditional chatbots that are keyword based, so you can actually ask a question and it understands the intent of your question. Voice assistance now AI-powered voice assistance and chatbots can provide 24/7 support, and that really just makes it more convenient for customers, enabling more self service.
There's people in our home offices that are ready and willing to take the call, but some folks may have a different shift or may be working different hours and it may not be convenient for them to call during those hours. So, this opens up more opportunities. AI now can translate on the fly. So if we're having a conversation and we don't speak the same language, AI in the middle could be the one that's translating to enable multilingual support, and this allows carriers to serve diverse populations, potentially underserved populations.
I find it very interesting that AI can oftentimes explain things to customers very well, and they're infinitely patient, so walking a customer through an explanation of benefits like 'here's what each field means, here's what your own statement is trying to convey.' So there's just different ways that, at a service level, AI can enable customers to have more options, more convenience and more help.
I also think about the fact that AI can enable us to help the customer in new ways. So potentially, think about AI-guided purchase journeys. If you have a policy, you need to talk to somebody, or need to find information about your policy. Or if you don't have a policy yet? What if you're just thinking about your needs and you need to be able to toss ideas around with somebody? There are always our producers or advisors, they're always there and willing to provide that guidance. Some people want to be able to just talk to technology and get that guidance as well. So, this opens up another avenue where customers could potentially go through an AI-guided journey where they have a dialog about their needs. The AI might ask them questions like 'what are your goals?' and then they could get some personalized product recommendations. And the AI could always ask, 'Would you like to have a conversation to pursue this further with someone who is licensed to sell insurance in your state?' Those are things that also increase convenience and enable customers to have a different experience.
And then there's the whole back office operation side. We can shorten the amount of time it takes to process claims and to underwrite, and that just makes everything more convenient for customers, for producers, even for beneficiaries. It's really all about what more could we do for the customer that we've been limited to up to this point? And how do we capitalize on this new pathway to serve them better and meet them where they're at?
How does agentic AI fit in and where is it useful?
Agentic AI is basically when we think about multiple agents. And agents, traditionally, in technical definitions, are a large language model, plus tool use, plus memory. So when we build something that works in our back office, usually we're thinking about an agentic workflow. That means many agents that are really specialists, work on specific tasks and then collaborate with each other to carry out what the user wants them to carry out. So some agents might be pulling information from systems. Other agents may be pulling information from documents. Other agents may be making certain determinations, gathering data, and they work together, just like people work together in a team, and they collaborate to produce an outcome.
Each indicator right now is very good at automating things that are structured, that are repetitive, that are rules based, where there's like, a clear right answer, and there are definitely aspects of insurance back office operations that fit this bill, you think about in claims, being able to detect what documents are missing, proactively requesting that, being able to route claims to the right adjuster and keeping the customer updated on progress so the customer doesn't have to call in to try to find out what's going on with their claim. Think about underwriting, you can have agents gathering information from multiple sources, maybe flagging things that are missing or that could be inconsistent, and preparing a kind of a brief for the person, the human underwriter, to review. That frees that human underwriter out to focus more on the risk evaluation, instead of digging around in systems to try to find information.
I think about sales too, in the producers and advisors world, being able to look up how products work. Asking the chatbot or being able to have AI taking notes and then potentially following up with customers when an email needs to be sent out or a reminder needs to be sent to the agent that frees up agents from can we focus on a lot of administrative work to be able to focus on working with customers and protect more lives.
Those are the types of things that agentic AI is very good at, structured, repetitive, rules-based tasks. Unfortunately, it can't yet, like negotiate things for us, go to meetings for us, and do all those things that really require tacit knowledge and also some amount of awareness of my environment, right? Who are all the people that I am working with? It's not yet at that level, but that doesn't mean we can't use it already, and there are definitely aspects to back office operations where it fits as well as even customer-facing frontline operations.
And we talked about some of those already in terms of agents and chatbots and things that answer questions for customers to serve them and enable convenience.
