Vineet Bansal, chief information and technology officer of The Mutual Group, is leading the insurtech's application of AI to its technology and operations support for mutual insurance companies. Joining the company last March, Bansal began these efforts last fall. Digital Insurance spoke with Bansal about what The Mutual Group developed for its clients, how mutual insurance companies are responding to the challenges of AI, and AI's potential for the future.
This is the first of
How did The Mutual Group turn to AI and begin using the technology?
In my fourth month with the company, we realized we had to start to do something in AI. We know AI will be a game changer. Mutuals are struggling in particular because they don't have talent, they have core issues, and they don't have capital to invest. We are in a very unique position to help them. We started with a sizable investment in the AI space. We hired people and partnered with other companies to begin the journey.
The sweet spot for us is mutuals that are between $200 million and $700 million in premiums, that do not have that ability on their own, where we could help drive it and be the thought partner. Our first use cases were claims and underwriting.
What has The Mutual Group developed with AI?
In claims, in the first notice of loss (FNOL) process, there are so many associated documents. AI is wonderful at going through all of it, and creating a summary about the claim. AI supports automated routing of emailed FNOLs to the right claims queues to create claims records. So claims adjusters are not focused on data entry, they're focused on actually working on the claims.
This FNOL processing is in place for our workers comp and property product lines. Our property liability portal is on track to launch in March. Beyond AI automation, we will work on human judgment aspects of AI, but those use cases are yet to start.
In underwriting, in commercial lines, agents are sending submissions as many emails with lot of attachments. They do use ACORD forms, but ACORD forms don't necessarily standardize how to submit the information. We do have online submission on our small business portals.
We are using generative AI capability to review email submissions, analyze forms, apply guidelines, and then save documents in our enterprise content management system. We also then create accounts and submissions in our policy administration systems. Also, we are about to make all of this available for commercial package products, which include property, liability, crime and inland marine coverages.
What principles do The Mutual Group and its clients follow for deploying AI?
The first is cost, undeniably. We can have fewer people do things but with higher quality and lower risk. AI hallucinates and we have to be cautious about it. AI is not plagued by human errors or potential human errors. So that's lower cost and better quality, but then speed is an issue. Can we actually quote a submission in half the time? We have an AI assistant doing things that otherwise take a lot of human hours, like improving the time to quote and the submission-to-quote ratio, because if we can quote quicker, we can quote more. Some quotes took 30 or 60 days. Agents don't want to wait that long. That's another benefit we are seeing.
We don't see AI replacing the intelligence of an underwriter, but we see AI assisting the underwriter to not do things that are repetitive in nature – go out to this website, get this information, go to the system to get that information. How can we create a single pane of glass so an underwriter can be more efficient and can focus on managing the risk better?
