I was recently interviewed for an article (more to come on that in the future) focused on how artificial intelligence could be used in insurance. This was a great conversation and centered around a few key questions that I think are important as we continue to explore how AI in insurance will impact our industry.
I am a big believer that AI will only enhance what we will be able to do for our clients. You won’t hear any doomsday predictions out of many about how AI is going to replace insurance agents. Now, insurance agents using AI may replace insurance agents, but it isn’t doom and gloom for all of us. The bottom line is that AI is an incredibly powerful tool that, when deployed in the right areas, can allow us to significantly increase the value we provide to our clients.
I thought it would be fun to answer a few of the questions the interviewer asked in more detail here:
How do you use AI in your day-to-day processes right now when it comes to sales and marketing? What about business management and agency operations?
This is an interesting question because AI in insurance is still very new. We’re experimenting. There are many use cases as it relates to drafting communication to clients, underwriters, claims reps, etc. If you are struggling to explain a situation to an underwriter and get the result you want, pop the info into ChatGPT and ask it to help you draft the email.
There is a huge application in sales and marketing. I don’t recommend having AI write your blog posts entirely at this point (see more on having a human in the loop below), however it does a fantastic job of helping you brainstorm ideas and outlines for blog posts and content. I have had ChatGPT write email marketing sequences that I can edit and make my own before putting them into an automated campaign.
Role-playing is another awesome application. Use AI to give you the top objections you may get from your client and the best ways to handle those objections. You can have it give you entire sales scenarios to practice and get better at certain parts of your sales process.
CJ Hutsenpiller, @cjhuts on Twitter, is doing a ton of great stuff using plugins and Zapier to connect different tools and create automated processes that incorporate AI. He has even created a course you can buy to understand many of the applications you can use to optimize processes in your agency. I certainly recommend following him. Coincidentally, he also has some awesome dad jokes.
Are you using any AI tools to help with policy personalization or risk mitigation?
YES! This is one of the most exciting areas we are exploring with AI. Currently, we are building AI into LAUNCH to help with a few specific areas. The first two areas center around supporting an inexperienced producer or service person to better understand the impact certain risks can have on a client.
When we go through an assessment, we need to ask the client great questions and know how those questions could impact them. AI is a great tool to help the producer better understand these impacts.
The next two areas are all around analyzing the data that we collect from risk assessments. AI is great at taking a lot of data and identifying trends in that data. We are having it identify the top risks that the client should be aware of based on how they answered the questions.
Once that analysis is done, we are having the AI recommend the best risk management solutions to manage those risks. This will allow a risk advisor to have an “AI Risk Assistant” that will do all the heavy lifting related to the data analysis. This will level up any producer into a risk advisor in no time.
What are the risks with incorporating AI into processes?
While I do believe there are mostly great applications to AI in insurance, there is one particular area that I think could be a risk for agencies to be aware of. I definitely believe that we still want to have a human in the loop. Completely automating processes using AI is still a bit risky. You have to remember that the AI is spitting back answers that it believes are correct, whether or not they are actually right.
If we don’t have a human involved, we risk this information going out to a client or prospect. This doesn’t have to be a major intervention by a person, but we need to review the information and give our stamp of approval.
I like to think of AI as an assistant or a junior level team member. That assistant can do a lot of the leg work researching a topic, drafting communication or many other things, but the senior level person needs to review and give approval before the information goes out to a client. It’s the same with AI, and we want to build in similar processes.
This was a fun conversation to have. I love exploring the applications of AI, and our conversation really only scratches the surface.
Stay tuned for information about the article coming out soon!