How Big Data and AI Are Reshaping Insurance Oversight

Pete Miller [00:38]: Insurance regulators play a key role in some of the most pressing issues facing the insurance sector today, from the accelerating pace of technological change to the ethical obligations that come with it.

The Institute’s Griffith Insurance Education Foundation serves public policymakers by providing nonpartisan resources and programs designed to enhance understanding of insurance and risk management. Given that mission, Frank Tomasello, Executive Director of the Griffith Foundation, is our guest host for today’s episode with Wisconsin Insurance Commissioner Nathan Houdek.

Few people understand the balance between insurance innovation and regulatory oversight better than Frank. I’ll hand it over to him now….

Frank Tomasello [01:23]: Thank you, Pete. It’s a pleasure to guest host this episode of the Predict and Prevent podcast. Public policymakers play a significant role in advancing efforts to predict and prevent loss. I sat down with the Honorable Nathan Houdek, Commissioner of Insurance for the State of Wisconsin and a prominent member of the National Association of Insurance Commissioners, where he serves as chair of both the Financial Condition Committee and the Big Data and Artificial Intelligence Working Group. Commissioner Houdek shared his views on the promise of innovation as a tool to better predict, prevent, and mitigate loss, and on the importance of fostering that innovation within the insurance ecosystem while ensuring that consumers are protected.

Frank Tomasello [02:08]: Commissioner Houdek, welcome to our microphones.

Nathan Houdek [02:11]: Thanks, Frank. I really appreciate the invitation and the opportunity to be here with you today.

Frank Tomasello [02:16]: The pleasure’s all ours. Commissioner, to begin our conversation, big picture looking at the RMI ecosystem today compared to five years ago, how would you characterize the changes that innovation and technological advances are enabling? And are we in the initial stages of transformation here or are we further along than most folks might think?

Nathan Houdek [02:37]: That’s a great question, Frank. So you know we’re always seeing innovation and new technological adoption in the insurance industry. That’s taken place forever. But I think over the last five to seven, 10 years, that pace of adoption and change has really accelerated. I think coming out of the COVID pandemic, you know, we saw a number of changes on the health insurance front with adoption of telemedicine and a lot more embrace of kind of that form of service delivery. You know, looking at accelerated underwriting and life insurance, right? That had been around for a while, but when you can’t go and get lab work done or get a medical exam in person, you saw a lot more reliance and kind of acceptance of utilizing accelerated underwriting, which you know I think it’s that’s a great advancement because it benefits people to be able to purchase life insurance faster and easier.

And then you know in the property and casualty space, we saw obviously with driving patterns being so disrupted with the pandemic, you saw insurers really starting to analyze data, driving pattern data in a different way, in a much more granular way than they had before. So I think you know you saw a number of changes come out of that. And then from that time with new sensors and Internet of Thing adoptions and obviously now AI and big data, we’re seeing you know all of this innovation just starting to take off at hyperspeed. And in some ways it’s building on what had been done before, but in other ways it’s taking us in very new directions.

Frank Tomasello [04:27]: Thank you, Commissioner. This move from an approach of focus on detection, a focus on repairing loss on the back end, to instead predicting and preventing all of this on the front end represents a significant paradigm shift. In practice, you touched upon this a little bit just a moment ago, but what has that looked like? Are there some specific concrete examples of that paradigm shift that strike you as particularly important in recent years. We’d appreciate your thoughts on that.

Nathan Houdek [04:58]: I think one area where we’re really seeing that take place is in the property insurance space, right? We know we’re seeing more frequent and severe weather events, the California wildfires. I mean, here in Wisconsin, just last month we had more tornadoes in one week than we typically have on average in a given year. So, you know, whether it’s whether it’s hurricanes or wildfires or severe convective storms, you know, we know we’re seeing more severe weather events. And so what that’s really kind of reinforced to your point is the importance of investing in risk and loss mitigation kind of tools, if you will. And in particular that’s really happening in the property space where, you know, I mentioned drones earlier. We’re seeing obviously insurers utilize drones to identify vulnerabilities and weaknesses, to try to make sure they’re preventing loss in the event of a severe storm. We’re seeing adoption of you know sensors in both homes and commercial buildings to try to you know detect if there might be a leak that could lead to a flood or some type of electrical shock that could lead to a fire.

And then you know where this is all really coming together in a bigger picture manner is the adoption of these kind of strengthening home programs in states across the country. Alabama was the first state to adopt this type of program. I think it was back in around 2011 called the Strengthen Alabama Homes Program. And since that time, you’ve seen a number of states, especially in recent years, putting forward these types of programs to provide financial incentives for homeowners primarily to invest in a higher building standard so that when they we are getting hit with more severe storms, their homes will be protected. And so that’s where I think you’re really seeing public policymakers, regulators, the industry, consumers really coming together to try to get on the front end of this and invest more in risk and loss mitigation.

Frank Tomasello [07:12]: Commissioner, you mentioned the Strengthen Alabama program. Can you talk a little bit more about that? Is that program one that has served as a model for other jurisdictions? What does that look like today? And is that sort of from your perspective perhaps an example of the greatest alignment we’ve seen within the insurance ecosystem around innovation and balancing that innovation with regulatory considerations? Appreciate your thoughts on that.

Nathan Houdek [07:40]: Yeah, I mean that program has really served as a model. And as I mentioned, you’ve seen a number of other states adopt programs that are maybe not identical but very similar, right? In terms of providing some type of financial incentive to homeowners to then build to a higher building standard. And you know, I think that is probably the best example of kind of all the players in the insurance ecosystem working together cooperatively, collaboratively, to really  accomplish some goals that are beneficial for everyone, right? we can’t stop, you know, the more severe wildfires or the earthquakes or tornadoes or whatever it might be, the hail storm that hit Madison, Wisconsin, here a few weeks ago. Like that is still going to be happening. But how can we all work together to try to prevent mitigate the loss and just build stronger homes.

So, you know, when you look at it, how it’s structured with the Institute for Building and Home Safety, which the insurance industry supports and is, you know, obviously a big backer of, that has really served as a model for setting higher building standards. The problem is it often costs more for people to build to that higher standard. So that’s where these types of programs come into play to provide financing for people, grants or tax credits, whatever it might be, so that people can build, the homeowners can build to this higher IBHS standard called a FORTIFIED standard to have stronger homes. And so obviously that benefits the homeowners, that benefits obviously insurers, that benefits from a regulatory standpoint. You know, we want our markets to remain competitive. We want people to be able to access affordable coverage and so mitigating loss helps with that as well. So, it really is a win-win-win program and it’s something that we continue to see more interest and more engagement around those types of programs.

Frank Tomasello [09:43]: Thank you, Commissioner. Want to shift gears a bit. You spoke a bit earlier about big data and AI, certainly very powerful tools that are creating significant opportunities to better predict, prevent loss, opportunities as well for mitigation in certain respects. and some might say that with those opportunities come significant obligations as well. I’d appreciate your thoughts on the ethical imperative there.

Nathan Houdek [10:08]: Yeah, well so back in 2020, the NAIC adopted the principles on the use of artificial intelligence. And if you look at those, there are a number of principles that were adopted, but the first two I believe relate to the fair and ethical use of AI. So, you know, from an insurance regulatory standpoint, you know, we know the ethical use of AI is really imperative. I’m sure we’ll talk more about it. Just the kind of benefits and other work that we’re doing around AI.

But you know, in general, we support as regulators, we support innovation. You know, we support technological adoption and advancement. That’s good for the industry, it’s good for consumers, it’s good for society overall. But we also need to make sure whatever that advancement or that new technology is that it’s being adopted in a way where consumers are protected, where it’s being used in compliance with all applicable laws and regulations, and all the considerations about how it’s being used in a fair and ethical way are taken into consideration as well.

Frank Tomasello [11:16]: Commissioner, as a follow-on, can we talk a little bit about the NAIC’s AI system evaluation tool pilot? Is it fair to think about that as a tool that sets boundaries rather than barriers? And if we maybe step back a pace or so, some of our listeners may be unfamiliar with that pilot. Can you help us to understand what that pilot tool is all about, why it’s been introduced by the NAIC, and how is that pilot operating?

Nathan Houdek [11:43]: Yeah, it’s a great question. It’s one of the big initiatives that the NAIC is undertaking this year. So I’ll back up a few steps. I like to talk about how as insurance regulators we’ve been trying to take a reasonable and measured approach to how we are regulating AI. So I mentioned in 2020 we had adopted the AI principles. Following that, we did a series of data calls, kind of focused on forming individual lines. So we did an auto, homeowners, life and health data call for each of those to understand how insurers are using AI in those four product lines. And that that information, the data that we collected helped us understand how the industry is using AI and helped us think about what are the next steps we should be considering in terms of potential additional regulation. There was some discussion at that time about should we move forward with developing a new AI model law or a new AI model regulation. But it was decided at that time that we didn’t feel we were ready for doing that.

So, what we did is through the NAIC we developed the AI model governance bulletin. And basically that bulletin kind of lays out how regulators are kind of thinking about insurers using AI and using it in a way where they are complying with all the laws and regulations on the books. So just basically taking a look at OK, what relevant and applicable consumer protection laws, unfair discrimination, etc., are currently in place. And now, as insurers are using AI, how do they how should how should insurers think about the use of AI as it relates to those existing laws. And that’s really what the model bulletin set out. And we’ve had about, I think, 25, 26 states have adopted that bulletin.

So, getting back to your main question, we’ve now moved on to kind of the next phase in our multi-step AI regulatory kind of journey, and that being the development of an AI system evaluation tool. And really what this tool does is allow .. it’s developing a kind of multi-part questionnaire, if you will, that again allows regulators to really understand how insurers are using AI. And it’s laid out in a manner where it’s very risk focused. So, you know, there might be many different use cases for a particular insurer. From our standpoint as a regulator, we’re less concerned about maybe the administrative uses, back office uses, but more concerned about where is there potentially consumer a direct consumer impact or direct consumer harm if it’s being adopted incorrectly or improperly. And so the tool lays out a formalized way for regulators to ask a series of questions and really drill down and understand how insurers are using AI and where should we, as regulators, be taking a closer look at that use? So the tool was essentially exposed earlier this year in its current form. And now over the summer, we’re going through a multi-month pilot process. So there’s 12 states that are piloting the tool. Then, based on this pilot phase and the feedback that we received from the pilot we will be refining the tool and re-exposing another version in the fall with the goal of finalizing the tool at the fall national meeting later this year, and then having that be available for states to use on a voluntary basis beginning next year as they are examining insurers use of AI.

Frank Tomasello [15:37]: Appreciate that overview. Thank you for that. To follow up, can you share anything about progress thus far with the pilot? How are things going in the early stages or is it too soon to really comment on that?

Nathan Houdek [15:49]: So we’ll be having a public meeting of the Big Data and AI working group in a couple weeks. And at that meeting, we’ll be providing an update about what we’ve been learning through the pilot phase to date. The 12 states that are in the pilot have been meeting regularly, in regulator-only settings to discuss company confidential information and share what we’re learning, what things are going well, what things might not be going well. Again with the idea so that we take this this kind of information that we’re gathering and feedback that we’re gathering during this pilot phase so that we can refine the tool later this year to come up with a finished product. So there’s a lot of kind of internal discussion happening as we’re working through the pilot and we will be providing kind of regular public updates at a meeting next month and then the summer national meeting which will be taking place later this summer.

Frank Tomasello [16:47]: This notion of balancing innovation and regulation is an ongoing effort within the insurance ecosystem, and it has been. Certain observers may take the position that regulatory oversight of AI may slow innovation, may raise costs, may impact consumers negatively. Others may suggest that you know responsible scaling and rapid innovation are not mutually exclusive. I’m curious to get your views there and if you could sort of help us explore the tension there a little bit, we’d appreciate it.

Nathan Houdek [17:20]: Well, as you mentioned, there there’s always some degree of tension, right, between innovation, new technology adoption, and regulation and oversight. So obviously we all know there’s a much bigger conversation happening around AI and AI regulation more broadly. but viewing it, you know, just through the lens of insurance regulation, you know, again, as I mentioned, we really are trying to take kind of a reasoned and measurable approach to making sure that we’re allowing companies to adopt and innovate, use AI in ways that hopefully help and benefit consumers, right? Speed up the insurance lifecycle, make it easier to purchase, make it easier to file and process a claim, you know, all the different kind of points along the life cycle where there can be friction or it could can be kind of difficult for a consumer, I think there’s probably potential AI use cases that could help along that that journey. So that’s a good thing.

And you know, we want to see insurers be innovative, but as regulators, we still also have a job to do in terms of making sure that consumers are protected through that process, which is obviously our kind of our North Star, our fundamental kind of objective as insurers or as regulators, excuse me. And so that’s where rather than taking an approach where we would be inhibiting any new adoption or potential use case by an insurer, we’re really trying to take an approach of making sure as insurers are adopting this new technology and these new tools that they’re doing it in a way that all laws and regulations are being complied with. And that’s really, you know, where we are right now in terms of kind of our regulatory approach.

Frank Tomasello [19:15]: Commissioner, you mentioned consumer protection and on the notion of consumers, I wonder if we might talk a little bit about consumers’ roles within the insurance ecosystem. From where you sit, are or will insureds become more active partners in risk management as predictive models evolve? Will they help inform insureds so that so that they can engage in preventative action? What might that look like? What are you seeing and what do you foresee?

Nathan Houdek [19:43]: Well, it’s a good question. And I do think, you know, going back to what we were talking about a little earlier with the strengthening homes program, you know, I think that’s a prime example where you are seeing consumers or in that case, you know, homeowners playing a more active role, being asked to play a more active role in terms of actually thinking about what steps they can take, what investments they can make to mitigate loss.

You know, it is it is kind of a different risk management mindset as a homeowner, rather than you just buy a home, you buy your insurance policy, and that’s it. And if you know, you have a loss, you file a claim. It is a different mindset in terms of OK, I’m a homeowner getting my roof replaced. Should I think about getting a stronger roof that’s built to a IBHS standard, even though it may not be required where I live. Should I think about getting stronger windows and better siding? And you know, there’s a lot of wildfire mitigation work that’s being done, as well. And I know out in California that there’s at least one kind of community that’s been built to a higher building code standard to mitigate wildfire loss. So it’s a it very much is a different mindset for homeowners and I think you’re seeing something similar with health insurance, you know, where with the increased use of wearable technology and more people have, you know, data about themselves, their heart rate and how they’re sleeping, and all this other information about, you know, their biometrics. Yeah, I think there’s an encouragement for people to kind of use that information and be thoughtful to keep themselves healthy and try to make sure that they, you know, don’t have a major health event. So, there is a different mindset and a lot of this is being supported by and driven by more data, right? Just more data overall, but then the ability to analyze that data through AI and machine learning, etc., in ways that we just didn’t have the ability to do before.

And so through like the NAIC Natural Catastrophe Center of Excellence, you know, we have the ability to, as regulators, we might, you know, internally in our department, we might not have data analytic capabilities at a very granular level, but through the Center of Excellence, we can be supported and be able to do data analysis in a way that we couldn’t before, which then allows us to identify potential areas in our markets where we might have affordability or accessibility issues for homeowners. So, you know, I know I’m kind of jumping around, but there are a lot of different a lot of different uses, and I really do think for consumers it is a different mindset. And leveraging data, leveraging the resources that are out there to be more active in risk management and loss mitigation is something that people frankly need to be thinking about more.

Frank Tomasello [22:49]: Thank you, Commissioner. As we look ahead and we think about the promise of innovation through big data, through AI, other technological advances, I’m curious. If you look out five, seven years from now, how do you foresee the various players within the insurance ecosystem, specifically public policymakers, members of industry, consumers, how do you foresee those folks coming together to advance risk management by better predicting and preventing loss, and/or engaging in in steps to mitigate. What do you see down the road a few steps?

Nathan Houdek [23:22]: Think there is more of an understanding and acknowledgement of all of the players that you mentioned in the insurance ecosystem really working together. You know, because at the end of the day, we want people to have the right insurance coverage, whether that’s homeowners, auto, life, health, you know, it’s important people have the right coverage. We want it to be affordable. We want it to obviously cover losses that people have in a way that it’s you know meaningful. But to achieve all of that, we also need to be thoughtful about some of the pressures that we’re facing. You know, whether those are more severe weather events that are occurring and the importance of thinking about risk management and loss mitigation as a homeowner, whether that is thinking about you know, the challenges we see in the health care, health insurance space with rising costs and an aging population. No one entity within that insurance ecosystem is going to be able to resolve these big challenges and these big issues that we’re that we’re facing. And so I think it really is not just important but necessary for regulators and public policy makers, and insurers, and consumers, and academics to all work together and leverage data, leverage the technology that we have, you know, come up with new innovative public-private partnerships and really make sure that we can continue to have accessible and affordable insurance across all lines.

Frank Tomasello [24:55]: Commissioner Houdek, for folks that want to learn a bit more about the role of public policymakers and efforts to better predict and prevent loss, any suggestions in terms of resources? Where can they go to learn more?

Nathan Houdek [25:09]: The National Association of Insurance Commissioners has a lot of resources on their website. There’s information about all the different committees and task forces and working groups and the various initiatives that insurance commissioners are working on through the NAIC. There’s a lot of FAQ documents and informational resources. There’s links to other partner organizations like the Center for Insurance Policy and Research, Natural Catastrophe Center of Excellence, and all the work that they’re doing and the publications that they’re putting out. So, they’re, you know, in a way I think the challenge is there’s so much information out there that it’s hard to know where to turn. But I would say that you know, especially for people in the industry or if you’re a consumer looking for kind of basic FAQ type information, the NAIC does have a lot of very helpful resources.

Frank Tomasello [26:02]: Thank you. Anything else you’d like to share with our audience, Commissioner?

Nathan Houdek [26:06]: I think just one observation that and you know, I kind of touched on this earlier, but I do think there’s a kind of a mindset shift that that people need to have as it relates to insurance and kind of personal risk management, if you will. I think you know, we’re seeing an increasingly complex world, we’re seeing you know a lot faster adoption of new technologies. We’re seeing new risks that are emerging that people have to think about in ways they didn’t before. And as part of all of that, it’s not just a matter of, well, I buy an insurance policy and that’s it. I think it is important to have the right insurance coverage, but it is also important to think about how can you as an individual or your family really think about kind of risk management in all aspects of your life.

Frank Tomasello [26:53]: Commissioner Nathan Houdek, thank you for visiting with us. We appreciate your sharing your insight with our audience.

Nathan Houdek [26:58]: Thanks, Frank. I really appreciate it.