00:03 - [music] [music] Hello everybody. Welcome to Insurance Business TV. We are here at Insurance Fest 2026 for another of our peer-to-peer interviews. And alongside me today, I have Mike Gulla who is the CEO and co-founder of Adaptive Insurance. We're going to talk about all things AI. Mike, is that good with you? >> Sounds great, man. Yeah, great to be here. >> Just tell me about the meaningful ways in which AI or indeed connected data is being used in the insurance industry right now, whether that's in

00:37 - underwriting or claims and what do you think about AI is still just hype. >> Yeah. You know, it's interesting. It's being used everywhere right now. You know, I think the interesting thing and it's come up on a couple of panels that that have been discussed here at insurance fest today around corporate governance, around AI and compliance. Everybody's using AI on their phones. They're using it on their laptops. So the industry, I think, is starting to move more towards this

01:01 - corporate governance approach of can we get everybody to use our approved version of AI because they're going to use it regardless, right? Whether whether they want them to or not. So it's better for them to take a more proactive approach to try to actually, you know, get after it now. Um, I think we've seen some impacts in the insurance space from AI. I think it's been more around the administrative tasks. You know, as you know, we work in a very regulated industry. So using AI to make

01:26 - decisions around claims or decisions around underwriting, I don't think that's fully happening yet. You have a lot of AI companies that pitch those capabilities that they can do. And I have no doubt that they could make those decisions. Um, but the risk to an insurance company around making an underwriting decision or claims decision based on an AI agent is very risky in a in a litigious environment that we work in in the US and in a regulated industry like we have with insurance. Um, but I am actually very excited to

01:55 - see how much the industry has taken a hold of AI and actually invested in those technologies and started to take a more proactive approach to it. It's been, you know, it's been good to see. It's been good to see. >> Let's talk a little bit more about that if we can because you said that the industry needs to be proactive, but there is this image that it's lagging behind in what areas do you think that it's struggling at the moment? >> Yeah. So, there's a data issue right

02:17 now, right? The industry traditionally has always been very reactive, right? So they don't usually pay attention to data until after a claim happens, right? That's typically when they look at what took place. They run their cat models. They figure out how much impact they had and they start to do these things. What we've started to see with AI is you've seen this impact that AI has now had to bring data to the front instead of from the back. Right? So now you have this proactive approach of what can I do with

02:41 - this massive gold mine of data that I'm sitting on that is the biggest asset for an AI engine model, whatever you want to look at, right? Like the insurance industry as a whole is sitting on probably more data than any other industry in the world. They have 50 plus years of claims data, underwriting data, customer data, all these different things, claims that they've paid and interesting things that they've learned about people's homes, businesses, and other things from claims. Um, so I think

03:06 - you're starting to see the industry look at how do we invest in understanding that a little better, >> very far behind other industries that don't have the same regulatory environment that we work in in financial services. So, it's allowed them to move a lot faster than the insurance industry. Um, but that's why I say I'm excited that the industry has actually taken to AI. Everybody talks about it. Everybody has somebody within their team, whether it's Geico or State Farm

03:30 - or All State all the way down to MGAs like us that have it at the forefront of a lot of the technology that we built. Um, but it's a positive approach and it's moving in a positive direction. >> I think I may know the answer to this based on what you've said so far. Well, I was going to say to you, what is perhaps the biggest barrier then for insurance to sort of integrate these tools given the fact that it has been such a risk averse industry? But you keep mentioning regulation. Is that

03:53 - it? >> Yeah, it's one piece, right? Regulation is always a piece that's going to impact the insurance industry, right? And that's not a negative thing. That's just something that has been around for a long time. Yeah. You have, you know, 51 plus different jurisdictions that you have to get regulatory filings in. Every regulator works a little bit different than the other depending on the state. So that is always going to be an issue. But I think the bigger piece that's

04:14 - been more interesting that I've seen in the industry over the last couple years specific to AI is the industry itself has never actually been afraid of investing in technologies. Right? This is why you have Guidewire and Uncorra and a lot of these policy admin systems. You have data teams at Vera Risk that do all kinds of cool things for the insurance industry and have for a long time. I think what is unique now that that you're going to start to see with AI and the piece that kind of

04:39 - scares the industry a little bit. They don't know that these models from AI are proven yet, right? That's the piece that's missing. So when does that gap actually close to a point that they actually trust the AI model and that there is claims data, analytics data? Actuaries are used to working on you know 10, 15, 20, 25 years of actuarial analysis going back in the past. So AI can do a lot to move that forward quite a bit. But the industry itself, I think, is still going to have a bit of time

05:11 - before they get comfortable with, do I actually trust that this is telling me something that I can make a financial bet on? Because if the loss ends up happening and it's a big one and I put all of my resource into using this specific model and that model was wrong, that is where the industry has been risk averse around new technology tools, around AI and machine learning models that have come out. And I don't disagree with them for that, right? I mean, there's a lot of things there that are

05:34 - that are still need to be worked out before that becomes, I think, the uh the forefront. >> Well, paint a picture for us, if you don't mind. What does an insurers tech stack look like 5 years from now? >> Can I tell you what I wish it would look like in 5 years? I would say I wish for the customer's benefit that they have to put in an address or they have to put in their name >> and the system does the rest of what is your risk. I would say that the other piece I would say in five years of what

05:59 - does this look like is and this is more of a hope and a dream and this is going to take a bunch of things to get us there but dynamic pricing >> that actually dynamically changes based on the risk profile y >> right a house in California has a risk that changes periodically throughout the year right at certain times the Santa Anas are blowing really heavy you know when those seasons are wildfire season rain season can we come up with a way that pricing can actually dynamically change at a rotation so that when

06:27 - somebody makes an update to their home, it has an immediate impact on their price, has an immediate impact on their premium. The same way gas prices constantly fluctuate by things that impact the world, right? Insurance pricing should actually change that way if it really worked. The real way that risk constantly evolves and changes, right? Most insurance companies underwrite a home one time at new business and then they have an inflation guard that mediocrely increases the, you know, the value of that property

06:52 - over time. data is getting to the point where in five years if somebody doesn't have the engine and doesn't have the technology tool to be able to optimize for those changes on a daily basis >> I think they're going to be way behind the eight-ball for where you know the industry is going because even with all the regulatory environment where it is I think you're going to start you you've already started to see regulators paying a lot closer attention and once the models start to prove that there can be

07:16 - a significant benefit to the policy holder >> the regulators are going to get more on board then the insurance companies will get more on board and they'll start to invest more. But 5 years from now, if you don't have the technology capabilities to leverage AI to its fullest extent, I mean, you couldn't even envision what AI is going to do in 5 years. If you look at where AI was 5 years ago to where AI is today and then try to predict where it is in 5 years, if you ask Google and SpaceX, I

07:44 - don't think they would even be able to answer that question where it's going. >> Yeah. So, [music]