Why AI Will Replace Risk Managers Who Don't Adapt | Kristen Bessette

Speaker A: M I think it's pretty clear that people with AI will replace people without AI. So given that, how do all these roles evolve? There will be entry level roles. They won't look like the same as they do now because just the same as when we brought in laptops or our parents aren't doing the same thing that we did when we started. So people with AI will replace that. So if you're not learning AI, you will self select out of your job because people will be better at it than you all of a sudden.

Speaker B: Hello, my name is Jason Reichel and you're listening to Risk Management Brick by Brick. I'm fascinated with people who are helping build and maintain the physical world around us. On each episode of this podcast we'll dive in with a risk manager, speak to them about how technology plays a role in this process. Hello, welcome back to uh, Brick by Brick. I'm your host Jason Reichel. Today we have Kristen Bassett, Chief Data Officer at Zurich North America. I always love talking to people from Zurich, they very data driven organization. You're uh, the king of data. So why don't you give me a little bit about your background, how you ended up in the seat you're at and what's in the forefront for 2026 and then we'll get into the questions.

Speaker A: So I've been with Zurich, uh, North America for about a year and a half. Prior to that I was at QBE North America, I was Chief Actuary Data and Analytics Officer there and then um, prior to that a lot of actuarial roles at Liberty Mutual. Across a bunch of views.

Speaker B: I'm so interested in the way that the actuarial model has to change given the new way we collect data. The new way, the better first party data we have and the more agentic third party inferences we can make, which is interesting to me. But before we begin with that, let's set the frame. Why are data and AI no longer supporting functions but core drivers to the risk strategy?

Speaker A: One of the really great things about AI is that it's really democratized the access to data. So now people have more ability to, to use technology to get things themselves self service all the information. And so what it's really doing is it's breaking down a bit. The silos that existed between teams and teams really have to partner together to deliver solutions now and bring those forward. So I think that's a real game changer in the way that we work.

Speaker B: Well, let's rewind it a little bit. In insurance, those silos were Needed because the data required deep specialization to be able to make sense of it. And now you're talking about this sort of generalist approach of pulling all those deep wells that insurance has into something that can actually be marketed and strategically viable to a business. Is that sort of the way you think about it?

Speaker A: I do think that requires some data preparation though.

Speaker B: Okay, let's talk about that. Let's talk about some data prep, because

Speaker A: I don't think you can just have everyone go out and use Copilot and answer every question they ever had with great accuracy. Right. So you still need expertise, you still need to understand your data, you need to understand your business, you need to understand the tools that you're using.

Speaker B: Yes.

Speaker A: Because to get the right outcomes, you have to have the right, um, inputs.

Speaker B: I guess the danger in the room that everybody's talking about when we talk about AI and data is if you can't understand that gap, you might be following, you know, phantom trails, depending on how your organization sets everything up.

Speaker A: Yeah, I think about the people who are, who are now building models, Right. It used to be really expert people. They understood bias, they understood data, they understood what the models were doing. They did all these testing around it. So the risk around that execution is a lot lower than when somebody who just is just learning how to use OpenAI tools goes in and builds a new model and they're like, this is terrific, I can do it. They can't necessarily evaluate, uh, whether it's the same product or not as they were able to build something. And so that's one of the risks around it.

Speaker B: Yeah. So let's talk about operating models, governance. Where does data create the most strategic value in insurance today? Underwriting, precision claims optimization. These are the things that we keep getting asked, like, how is the data changing? We talked about how it's changing the behavior, but how is it changing the entire approach to the insurance industry?

Speaker A: Well, I'm, um, a cdo, so you're not going to get me to say that data's only good in one place or another. Right. I think it has impact everywhere. But you know, if you think about the power that we have now, data, it's a very information heavy industry and there's a few different things you could go after. Right. You could say, how do I make this process more efficient? So how do I exchange information with all the third parties and put it in front of the right people and put it in our systems better? That's one way you can bring your data together and make a difference. Another way you can do it is how do I bring knowledge to decision makers? How do I get that information in front of them more quickly so they can make faster decisions, they can make better decisions.

Speaker B: You can disrupt it, you can disrupt those things.

Speaker A: Exactly. So those are the two ways I think about it. Those things exist in all of our different processes, right? So I think you can do that everywhere. It's just each process might have a different prioritization, each company might have a different prioritization of which one you go after.

Speaker B: I also find that often it's like AI usage for context aware individuals trying to improve their processes while leaving the um, sort of overall strategic to some more business systems. Right? How are you guys breaking down individual usage of AI uh within your organization and sort of corporate AI strategy?

Speaker A: And, and they do have to marry together, right? Because I think you're exactly right. I want everybody in the company to use AI and change their world because I think there's terrific power in little tiny transformation. Right? And so all of those people know their jobs better than I do. They can make a difference and do that every day. And I want all of that. But at the same time, if you're talking about how do I rethink what an underwriter does or a claims professional does and then have that be production ready to handle a lot of volume, to handle a lot of complex transactions, you need to have a more of a pro code approach to that and you need to sit with a business and redesign it with more expertise. And then there's a place in between which based on your organization and their maturity, you've transitioned between those two organizations.

Speaker B: Being from Silicon Valley, I know the best model for any data enablement is that you have data teams potentially and then you enable them, enable your company to understand that data and use it. But let's be honest, like a lot of times people would just get requests, the data team would develop the solution and then give the reporting out to the person. How are you going to, or what's the strategy for making sure that individuals within your team can actually train up to where the AI uh capabilities can be? What's that strategy look like? What are you recommending to other organizations?

Speaker A: I think from a training perspective there's a few different things. One is just knowledge, right? You have to get people know how to use the tools, understand what they're doing, understand what you roll out, what the outcomes are and how they evaluate

Speaker B: that mean finding a champion in each.

Speaker A: There's champions, there's formal training, there's on the Job training. There's so there's different ways there, but there's a lot of things culturally you have to overcome in that. Right. Because you really want to redesign processes, not just AI. A crappy process. Exactly right. You have to get also people thinking about how do I do this differently given that I have AI, what does this look like in the new world? And if I rewrote my process from scratch, what would I do? And that's a different sort of skill set that we need to build in the organization to kind of rethink and re explore and relearn.

Speaker B: Yeah, One big thing I'm seeing a lot of risk managers and the way that they're using AI and the data itself is restructuring. Okay. We understand it this way, this other team understands this way. So using the AI is more of a collaborative hub, transformational layer to get people aligned with more while they're doing the work versus more of like a presentation down approach that we used to take with data and do a big briefing and all of these kind of things.

Speaker A: There's a language barrier between teams. Right. Uh, particularly, you know, it and business or different parts of the business. How do you get everybody to speak the same language? AI can do that. And also if it's a big process, everybody has to be together to do it. And then you get a few experts with different backgrounds that are those translators. They could be your AI champions. They could be people like the actuaries you talked about who just know both sides a little better and could do that. And uh, that's really helpful to doing that.

Speaker B: One question I have about actuary because I just find it to be a fascinating practice. The insurance industry is based on that. Do you think we're going to see that really transform and evolve now that you can get more firsthand or quicker information back from AI tooling or machine m learning, all of these kind of technologies that are really picking up steam and are now being able to be managed. Are we going to have to look at how we do actuary work differently? What's your take on that?

Speaker A: Absolutely. And I'll give you a really dated example, but I think it's a good one. When you think of personal auto and decades ago machine learning algorithms went into personalized rate making. And uh, the whole industry talked about the use of credit and bias and all those things. But the end result of that was the involuntary market just went away because we understood the exposure and how to price for it. People were available.

Speaker B: Right.

Speaker A: Insurance in a way they never had before. And so now Fast forward to now. Better tools, more information, quicker access to that. What will we think of now that we can understand exposure? What will be the next thing that'll be market transformative because we understand the exposure, we understand how to rate for it, what products can we offer, what places can we reach? And so I think that's the type of power this has only scaled up from what's actually happened before.

Speaker B: One of the dirty secrets that I keep running into is we give tools to say a risk manager within a corporate environment, they think they know what their risk appetite is, they think they know what their exposure is. They run these through these models and they get a very different picture than what they believed their exposure was or these things. How does a risk manager from your perspective have that conversation when they were the ones accountable to it before and now they're at this turning point where the picture is much more clear.

Speaker A: So that's a good question. I think it's kind of like you have a hurricane barreling down and people say what's your exposure? And you just add up all your tiv. Right. Which is, which is not the right answer. And it will 100% freak people out.

Speaker B: Yeah, yeah.

Speaker A: So that will happen here. And then they'll have to be conversations like what does this actually mean? Is this really my exposure or is it something else? How do I get between these things? How do I think about this with the people they've been working with all along? Because there could be all kinds of causes for that. One could just be you're looking at numbers you've never seen before and you don't know what they are. It could also be that the model's not really accurate and you need to go back and refine the modeling to get a different result that's more in line with what you were trying to get. So those follow up conversations to understand what's coming out is really important in building that expertise and knowledge.

Speaker B: I think also training risk managers and more data practices like running more scenarios, things that they probably in school you're taught the risk registry, do this, do this, do this. And then in practice a lot of these people can't do that at scale or uh, there's too small of a teams and uh, putting those practices back in, in so they can see, okay, let's run different scenarios. Let's figure out how we can make you know, insurance buying more strategic, operational excellence more strategic, all the different elements, safety more strategic inside of like thinking about construction companies in their risk. How do we pull all those together to run scenarios so that we can really manage the risk in a more holistic approach.

Speaker A: And you have to give the models context, too. So when you're trying to do something, you'll get a better answer if you explain to the model who you are and what you're trying to do. I'm a risk manager. I'm trying to understand this specific thing and I'm looking for this information. Help me think about it and give me five different scenarios that gives you a better result up front than if you just ask it a question out of the blue and it'll give you something back and then you'll have to get there. Five further prompts. But it's a longer process.

Speaker B: One of the things that every time I post an episode where we talk a lot about AI, sort of the comments are always like, that's great, but we're talking, like, about a future 24 months away. The way that it's accelerating doesn't really seem that way. So what's some concrete wins in the last year and a half at Zurich that you guys have been able to put in place that you don't think 12 months ago would have been a reality that you can speak to?

Speaker A: There's any number of things, and it goes back to process and knowledge. So some of the things we talked about just bringing all that information in, bringing it in front of the underwriters really quickly, using, uh, visuals. We work with vendors to bring in visuals so the underwriters can see really quickly where a property has risk and where it doesn't. And the AI will highlight it so you don't have to, like, search it yourself. It'll just say, that's a spot. Looking at benchmarking, so we can go to our customers and say, here's where you're different than our other customers, or here's where we're better than our other competitors with real data. Looking at trends in our data, we can just sit there and ask our data questions and have it help people understand what's going on and what the

Speaker B: driver build that narrative case and build that narrative case.

Speaker A: And so I think there's a lot of great examples of what we've rolled out, how we're doing things better and faster than we've ever been doing. Like I said, it's changing all the time.

Speaker B: Let me ask this question about the intersection of technology and data. How do you make the decisions now when it's going to be, are we going to buy or are we going to build? How are you making that Decision differently than you were 12 or 14 months ago?

Speaker A: Well, I think, you know, in the beginning we thought we'd build everything. I think now the pace of change is so big, the vendors are offering so much that we do explore a lot with vendors, particularly if it's one of our major vendors or policy admin system. We use Guidewire. If they're building it in, Salesforce is building it in. Why would I then build something that does the same thing?

Speaker B: Right?

Speaker A: To do CRM, especially when you're getting

Speaker B: more value out of the platform that you already.

Speaker A: Right, so is it already built into something we're using? Is it offered by one of our hyperscaler partners, you know, on Microsoft or one of those folks? Is it actually a, uh, competitive differentiator? Because if it is, I want to build that myself. If it differentiates us in the marketplace, that'll be mine. Can somebody do it faster for me because I don't have capacity? Then I might go to a small startup just because they can get me in and do something quickly. Right. So there's all that type of trade off in that conversation. So we, we look at the demand and we evaluate it, the best way to go about it and then we decide how to build it.

Speaker B: This is a follow up question to that because if you're going to go the route now where you build some stuff, buy other stuff, how do you attract and retain top talent in the world of AI when it's such a critical change that this is where people probably should be focusing their career at this point in time? From my opinion, how are you able to attract them into the world of insurance? We're not known for being super sexy, right? How do we attract those talents, how do we keep those talents and how do we retain that into the. Not only build that talent up, but bring it into the organization.

Speaker A: I don't think anyone started out trying to be insurance. I mean, I didn't. Right. So it is a marketing problem we have. And I do think you can do anything in insurance technology. So we do try to market for it. I think though, at the end of the day, it's just like anything else. If you have the right culture and you're doing interesting work, you can get good people in that are excited about it. Right? So if you're doing work that's powering your business, it's making a difference. People feel engaged, they want to work there, you can get them in and keep, um, them.

Speaker B: One of the problems that some people have, you know, privately talked to me about is oh, okay. We work in this. I'll uh, take a carrier example. We work in this carrier. Three fourths of our resources will not train up to the level that we need to. But they're driving 75% of the revenue. Is that a fundamental existential threat or is that just bring new talent in under those people? Like how would you handle that problem within an organization?

Speaker A: I think it's pretty clear that people with AI will replace people without AI.

Speaker B: Okay, great.

Speaker A: So. So given that, how do all these roles evolve? There will be entry level roles. They won't look like the same as they do now because just the same as when we brought in laptops or our parents aren't doing the same thing that we did when we started. Right. So people with AI will replace that. So if you're not learning AI, you will self select out of your job because people will be better at it than you all of a sudden. And that's the reality, and that's your choice. If we give you all the resources that we can, and we say this is what you need to do, you can decide what you want to do with that and choose your own career path.

Speaker B: Let's talk about. You've worked across pricing, reserving, capital modeling. Where do leaders most underestimate interconnected risk? Like when you're working with all these data. Everybody likes to be transactional with data, right. They don't like telling that narrative story that data actually is, which is an ecosystem. Every lever kind of impacts every other lever. So how do you bring that together into an interconnected picture for your organization?

Speaker A: You know, it is tricky because you're going to have different people doing all of those things potentially. Right. So I think the reality is though, is that pricing risk becomes reserving risk, reserving risk and pricing become capital risk. And so they are not independent risks they used. It's a funnel, right? So if you miss on your pricing, particularly on a long tail line, the odds are much better you'll miss on your reserving too, because you just wrote different things or didn't hit the price points or the trends are different than you assumed, whatever the drivers are. But the reserving is reliant on the information it got when you wrote it to set up those initial reserves. And so if those assumptions are wrong, you're going to have reserving misses same thing on the capital side. So when you think about how things are changing now, the more information we can get in faster. So you can see that what you wrote is different than what you expected or what you thought the risk goes down.

Speaker B: Yeah, we're only talking about forecasting here versus we're talking about forecast actually. And then the faster you can reconcile those things.

Speaker A: That's right.

Speaker B: As a value differentiator.

Speaker A: As a value differentiator, that's great.

Speaker B: How should insurers think about balancing growth ambitions with capital discipline in this volatile market that we find yourself in 2026?

Speaker A: That's a tricky one. I mean that's always a balance, right? I, I, uh, think, you know, when you're thinking about it, growth trajectory comes with losses, right. So you do have to think in your capital structure how your growth, how risky that is. Is it new products, is it similar products? How does that look? Does that change your risk?

Speaker B: Is the insurance industry finding net new logos going down by expansion within existing relationships?

Speaker A: I think there's new demand. Like just think of data centers for example.

Speaker B: Right.

Speaker A: I mean it's still construction, but it's a whole new area that wasn't around. Right. And so that's a lot of demand that's, that's come in this way. And so when you think of new risks around that or cyber exploding or any of these things that have changed over time, new risks have different risks with them than personal auto. Right. So if you're a company that's going into newer areas, you just have to think of what that growth means from how much capital you want, what your risk appetite is around your, your capital buffers with that. Very few businesses are loss free entirely. So there's going to be losses as you grow your premium, then you're going to grow your reserves, which also adds to your reserve risk. Right. So running those scenarios that capital modeling exercise is important to do to understand those trade offs, are you doing them

Speaker B: more often than you were in the past?

Speaker A: I don't think so. Like that'll depend on the carrier. Right. I mean it just depends on uh,

Speaker B: what about product and service design? Has, have you seen that be impacted in the last 18 months for like really being able to find niche products or programs to put together to bring to market? Is that accelerating?

Speaker A: Yeah, I think companies that are listening to the need from their customers are trying to get product out quickly. You know, we got our data center product out very quickly into market. Yeah. Relative to maybe what we would have done a couple years ago. And so I think that that's a good example of how we were responding to what we were seeing and what people actually needed from a.

Speaker B: Did you spot those trends early on and were able to respond to them or were you Able to pull together a product quickly because of the data. Like was it the trend analysis? It was a both.

Speaker A: It was kind of both. But I mean we pulled it together very quickly. What, you know, what is the business saying you need? What do we have to execute on from the IT side? How do you get the pricing set for all those things? Right. It's a combined effort to get a product out and we just executed on uh, that really quickly into the market.

Speaker B: Amazing executives sitting within commercial businesses who great, let's call them business growth CEOs, CFOs are now in this new technology enabled growth world. I think the companies are going to win, are going to be technology enabled more than the historical relationship based ones. Hopefully you're using the technology to impact your relationships. What kind of advice would you give to the risk management listeners at home on how to have this kind of conversation, say with your executive team? Should they be pushing what's the hum on the ground when you're talking to commercial businesses or when your company's talking to commercial businesses? Where is the reality at right now?

Speaker A: I think most people are excited about the potential, but they want to see the reality. So the more you can actually bring things to people that show what you're doing or show it, it should be easier to have fact based conversations now about things. Right. So if you're a risk manager and you are worried about something, getting the information of a fact based conversation about it should be easier to do than it used to be. So that I think is a good place to start.

Speaker B: No, it's easier to get the facts than it was previously. Work on getting the facts right. Right. And then bring that to the table. I definitely think that because one thing I'm noticing is a lot more risk managers being elevated to executive level within their organizations as as that function really becomes a strategic lever within the organization. Let's play riskier. Too risky. How about that?

Speaker A: Sure.

Speaker B: Okay, great. Deploying AI models into production with weak governance risky. What is weak governance versus strong governance?

Speaker A: I think there's a couple pieces to it. There's a compliance piece. Follow the laws. Right. Follow the regulations.

Speaker B: Yeah.

Speaker A: Start or let's do that.

Speaker B: That's weak.

Speaker A: That's. If you're not doing that, that is weak. Right. But I think on top of it, you just have to make sure you understand what the models are doing right. You have a framework around ethical use, around data privacy and protection, around transparency, around outcomes. Make sure you're doing that.

Speaker B: Did you have to reinvent your data governance rules and all of those meanings with AI or did most of what you had previous apply?

Speaker A: Most of what we had applied. There are things that are different. So for example, with AI models you don't have hold out data sets that you then test outcomes again. Right. So the whole testing around bias or around what the model's doing is different. So that had to be specific. But largely we have a model governance policy that requires all the things that we use for AI.

Speaker B: So I think that's so important to say because so many people think that AI cannot produce results that are repetitive, but in the right governance they definitely can. Right. So a lot of people think it's a black box, but you can put these governance in so that you can really understand it. I think that's a really important message.

Speaker A: You can certainly make it happen.

Speaker B: That's the fear that people have. And the reality is probably as good as any data warehouse you ever had before done the right way. Right. It's just a uh, thinking up through the process.

Speaker A: Well and the other thing that people think about with their process is this isn't giving me 100% accuracy. I guarantee you the person doing it before was not 100%. 100%. And so there's a little expectation thing you have to work through and that's also a risk tolerance around what the model is doing. You know, some things are less risky, some things are more risky and you can kind of adjust your model and the amount of effort you put into getting those out outcomes aligned with the rest.

Speaker B: I think about that all the time. I live in San Francisco, so Waymo is an example. It's like, and I'm, I'm all for this, I don't want people get hate mail. But like if a Waymo hits something, it's the end of the world. But people are hitting stuff all of the time, once in a while. So this expectation of technology is not really. We have to accept the risk that comes with using this technology and learn how to manage that risk as a net new function. In my opinion, uh, and you have

Speaker A: to think about all risk is not the same. So if you're just using your version of a copilot tool to go in and read your email after you came back from vacation, that's fine. There's no risk in that. If you have a customer facing thing, maybe that's riskier and you wanna spend more time to make sure that the direction it gives is more accurate. You know, you hear bad stories about someone negotiating with uh, an online tool and they ended up buying A car for a dollar because AI sold it to them for that and they bought it for a dollar because that's what happened. So be aware of what the risk

Speaker B: is treating data quality as an IT issue rather than an enterprise issue. Riskier. Too risky.

Speaker A: That is risky because from an IT perspective, if the data is accurate, going from the front end system down to me, that's not an IT data quality problem because somebody is typing in the wrong data.

Speaker B: People are surprised by how often that happens.

Speaker A: Is a business owned problem. Now it's everybody's problem. Data quality is everyone's problem. But it has to be jointly owned. It's not a data governance team issue. It's not an IT issue. It's a collective issue that we get the right information to actually feed into the models and get the right answers out.

Speaker B: Assuming model fairness is solved after one review, that is also risky. How do you guys test for fairness? We talked about. Do you run the same data set? Is IT all regression testing?

Speaker A: Depends on the type of models. Like I said, in, you know, machine learning type models you can do hold out data and test. And yeah, there's different things you can do. When you're looking at AI has to be a little more outcome focused and you have to test the, uh, kind of the process.

Speaker B: Yeah. How close it is to the target

Speaker A: process around it and the steps around it. So we, we have a whole kind of framework around how we do that and deliver that and make sure that what we're looking at is giving us the desired outcome.

Speaker B: Considering that, you know, maybe 20 years ago in this industry, people didn't go to school and find themselves in this industry. They didn't set out going, hey, I want to be in insurance or I want to be in risk. But now that is. Now it is. Right now it is. It's a popular, that's a popular job. I really take that seriously. One of the things that we want to do with the podcast is make this exciting, let people know that this is an entire ecosystem. If you're a technologist, if you're a futurist, if you're a data person, there's rules within this industry for everybody. Right. What's a piece of advice you would give to those young people when they're starting their career that you think would be helpful for this day and time?

Speaker A: Well, like I said, people with AI will replace AI, so they should use the tools. And most colleges are doing it. My sister's a college professor, so she would suggest they're using too much AI, but I Think do that. But most importantly, like, if you are coming out of school and you want to go in, learn the business you're going into. So even if you're in a technical role, an actuary or IT or something, the role is not just about the numbers. It's about the customers and the products and the things that we actually do to actually provide the financial security to people to help their companies achieve, or their. Their own personal lives achieve what they want to do. And so the closer you are to understanding what it is we actually sell and how we do that, the more successful you'll be in your career, regardless of what it is.

Speaker B: That process mapping is a really good piece of advice because if you can do that, then you can see how you can plug yourself into the organization. Right. And I think no matter what day and age we live in, people create gaps. Right. And if you want to be a valuable person coming into your career, fill those gaps, not just do the job.

Speaker A: Understand the customer. Right. We're selling a product. So even if you're a person who's kind of far away from it, understand what it is that we're selling, understand the value that we're adding to that customer, why they're buying it, and then you can figure out how you can increase that value, and that's how you become successful.

Speaker B: Is there anything else that you wanted to mention that's going on with Zurich or anything that you're talking about at rims?

Speaker A: Uh, this year, I am on the innovation stage tomorrow, so that'll probably be after the fact, once this goes live. But talking about AI augments and underwriting,

Speaker B: augmented underwriting, I love that. That's such a great topic. Is that part of the overall strategy for Zurich? Talking about that and trying to bring attention to that in the industry? Because I think it's really fascinating.

Speaker A: Yeah, I think we do often talk about what we're doing in that space with thought leadership.

Speaker B: Uh, yeah, I love that that's part of your guys's thought leadership.

Speaker A: That's what we're doing. Yeah. But I think, you know, broadly, we have our strategic plan. We're executing on that, like I said, trying to serve our customers, rolling out these new products. So a lot of exciting things that we're doing and, you know, happy to talk about it.

Speaker B: Yeah. So we were talking about the actuarial models, and we're always fascinated about this because our customers are often impacted by how businesses look, how. How brokers, how carriers get to the pricing. And so one of the things they're always interested in is with all these new factors. One, how fast uh, is how product gets priced and how product gets rolled out, changing. And then what are some of the factors that risk managers could be aware of now to help them make the case at renewal time that they're actually running a good risk program that they could, they should be able to get something out of the changes that they're making within their. Because like a lot of organizations are investing a lot right now. Right. And so how can they actualize that?

Speaker A: Maybe I'll start with the second one first. I think if you're an organization, the more information you can bring forward to help people understand what you're doing with your exposure and how you're getting better, the better off you are. So I'll talk about that maybe just in the context of insurance, what insurance companies need to do to buy reinsurance, because that's our consumer experience. Right. So if we've made changes to our portfolio because we have better information and we're entering different classes and we're exiting different classes, that's a huge piece of information. The reinsurers would want to know if we've made changes to our pricing, our limits or our uh, structure somehow coverages. Those are huge pieces of information that the reinsurers would want to know. And so in the past that may have been hard to bring forward. Now it should be a lot easier to bring forward and again have that fact based conversation about this is what we look like and this is why it's better than we looked before. And you should reflect that in either, you know, the coverage you give us or the price you give us.

Speaker B: One of the tools that I talk to a lot of risk managers doing is doing a monthly summation of how your programs are maturing, what data do you have under that? And so waiting to renewal time and trying to pull that together in a two week period. Yeah. Have it be being generated, have conversations,

Speaker A: have metrics, have things that you're monitoring that you can say we put this into place and this is how we executed on this and this is where we're better. So if you're rolling out a safety program, show what impact it's having. Right. Don't just say we rolled it out, actually get some data and facts and bring that forward. And I think that can really help you. So even in uh, the construct of your current program, you can talk about things you're doing differently. And whatever I think for us, you know, on the carrier side, when you look at the way that we think about some of our pricing models and things. What I talked about a little bit was how do you uncover new things, how do you uncover trends, how do you cover information in the data? AI will help us do that. If you can get into your data, you can get more complexity surfaced earlier and then try to understand what's driving our own profit and loss and where we have gaps or we have blind spots, or we're uncovering industry trends we didn't understand before. And so those, then we can decide what we want to do with them in the context of our pricing models. So I think that can happen faster and it happens in both directions. So it's not just a negative for the customer. We can also say, hey, this is a segment that looks great to us. Here's an area where we're under, we're under investing in it, we're under investing in it. Let's go and grow here. And then this is an area where, where maybe we're over invested from what we see. And maybe this is a place we're going to manage our exposure more carefully. And so I think it works in both directions, but it's really about like, what is that data? What is changing in that data and how do you have that conversation?

Speaker B: Are you seeing those conversations really transform in real time? Are you seeing a more mature conversation start to emerge as companies pull this data together and can understand it?

Speaker A: I think it's early stages and it's going to be different in each dynamic. Right. Each customer is different. Each dynamic is different. You know, there's places where we'll go to the customer and say, hey, you know, we're actually seeing a lot of this type of loss and maybe you should incorporate that in your safety program. That's great information because that will drive their experience mods, that will drive a lot of the things that they're doing. Right. So. And it's better for all of us that we have a safer workplace. So insights like that are great for everyone. It's still a maturity thing on both sides in terms of how fast and, and how robust those are.

Speaker B: That's what they're really worried about. They want to know how to keep their insurance premiums down.

Speaker A: Yeah.

Speaker B: Well, thank you so much because one of the things that I'm excited about is you giving people confidence moving forward. What your guys's organization does. There's a lot that will follow suit on that. So I appreciate you sharing that information. Sharing here on the podcast, but also sharing in the stage with the broader thing. And I would love to have you back for a roundtable talking about that in the future, so I really do appreciate it.

Speaker A: Sounds great.

Speaker B: Thank you very much.

Speaker A: Thank you. Foreign.

Speaker B: Brick by Brick is brought to you by Trust Layer. Find out how Trust Layer manages risk so that people can build the physical world around us. Head over to Trust Layer IO and then make sure to subscribe to Risk Management Brick by Brick on Apple Podcasts, Spotify, or wherever you get your podcast. On behalf of the Trust Layer team, thank you for listening.