AI in the Insurance Industry

In this episode of The Reminger Report Podcast on Emerging Technologies, host Zach Pyers is joined by Ian Irish, a law clerk from Reminger’s Columbus office and a law student at Ohio State. 

Together, they explore the dual role of artificial intelligence in the insurance industry—how it’s being used to detect fraudulent claims and, conversely, how fraudsters are leveraging the same technology to deceive insurers. 

From deepfakes to ethical dilemmas, this conversation dives into the evolving landscape of AI-driven fraud and the tools being developed to fight back.

Transcript

REMINGER REPORT PODCAST ON EMERGING TECHNOLOGIES

ZBP      Zachary B. Pyers, Esq.
II           Ian Irish

 | ZBP | Welcome to another episode of the Reminger Report Podcast on Emerging Technologies. I’m Zach Pyers and I’m here today joined with one of the Law Clerks from our Columbus, Ohio Office, Ian Irish. Ian, thanks for joining us and welcome to the program.
| II | Yeah, thank you. I’m glad to be here. I’m a law student; I’m entering my final year of law school this upcoming fall and I go to Ohio State.
| ZBP | Great. Today we’re going to be talking about the two faces of Artificial Intelligence in Insurance Fraud. So, you know, as we are looking, talking about Artificial Intelligence, which has obviously been a very hot topic over the last couple of years. Explain to us, kind of a snapshot view of how this is working as it relates to insurance fraud both in the positive and the negative.
| II | Yeah, so at the end of the day, AI is a tool, right? Like any other tool it’s a matter of how you apply it and whose hands it’s in. So, on one hand we can have insurance companies utilizing AI, utilizing generative AI on their end to help them detect insurance fraud or claims that don’t quite seem right. On the other hand, you can have scammers who are doing the same thing on their end but to carry out fraud, including insurance fraud.
| ZBP | So, when we look at this right, as a tool, how is AI currently being used to detect and to prevent some of this insurance fraud?
| II | I think as you know, this technology is pretty new and I don’t want to give the impression that this is being used widespread by insurance companies, but some are definitely using it and the ideas that generative AI is good at spotting patterns and it does that really quickly, really efficiently because it’s a computer of course. So, insurance companies can utilize these AI models to determine if an insurance claim checks out, if there’s any red flags that suggest fraud or that the claim is overinflated or something like that.
| ZBP | I remember years and years ago I had a friend – this was well over a decade, it was probably closer to 15 years ago, who was a data scientist who actually worked for a very, very large insurance company and I remember having lunch with him and you know, he was a very interesting guy. He told me – and again, this was almost 15+ years ago, he said insurance companies have so much data at the time – he goes, we don’t even know what to do with it all. And again, this was like 15 years ago and what I think you probably are seeing and some of those – they’ve been using that data that they’ve been collecting for years to help to kind of trigger or look at some of those red flags that you’re talking about that AI is able to pick up on probably faster than a human would be because its able to review more data sets faster.
| II | Yeah, I mean I’m sure they’ve been accumulating this data for years and years and years and now finally they have, you know, these models that they can feed it into and put it to good use in the way that, you know, people can’t do anything with that, but computers can.
| ZBP | Right. Now, you know, we talk about this – and I’ll be candid, when we talk about, you know, the fraud measure, the fraud techniques, I sometimes feel like I’m the last to know, right? Because I’m not creating fraudulent schemes of my own and so when I read about these, I’m always kind of a little surprised. How are fraudsters, how are, you know, people creating these fraudulent claims using AI?
| II | Well, you know, I think the word that comes to mind, and we hear it a lot – kind of a buzz word is ‘deepfake’ right? This audio-visual representation that imitates an individual, it can imitate their likeness and their voice, and you know, these scammers can use these fake people to do things like change a policy beneficiary or secure a policy loan by imitating the policy holder. And I think a lot of the concern comes from how quickly these models have improved in the past 5 years and you know, 5 years from now, who knows how convincing they’ll be; how easy it will be for, you know, Joe Schmo to utilize them.
| ZBP | Now, you talked about some of the deepfakes and some of the risks that we run, and we’ve seen this play out in several industries before but what makes the deepfakes particularly so dangerous in the context of an insurance claim?
| II | Yeah, I mean I think about what we traditionally think of as like a scammer, right? Usually, you get an email message, a text message, has probably some spelling mistakes. You know, if you’re reasonably comfortable with that stuff, you can pretty easily sauce out hey, this is fake but now we’re talking about things that to an uncritical eye look no different than you know, if the policy holder had actually contacted the insurance agency. You know, these agencies already have lots of claims they have to sort through, it’s a fast-moving industry and now, you know, take convincing, fake, low-cost scams thrown into the mix and that’s a risky thing.
| ZBP | Are there tools that are being developed, you know, especially with the deepfakes, to kind of detect and kind of counteract this – I don’t want to say scheme but this problem that some of the insurance companies are facing.
| II | Yeah, and I think it depends a lot on the quality of the scam. You know, there’s some kind of deepfakes that you or I could take a look at them and tell pretty easily, you know, this isn’t real – maybe the lips don’t sync up quite right to the audio, maybe there’s some weird clipping with the audio, maybe the background, the lighting and the texts just don’t look right. You know, if we go on, there’s a lot of AI on Facebook, right? A lot of AI Facebook posts that become very popular and when I look at something like that as someone who’s maybe grown up a little bit more with this stuff – I can tell it’s fake. But some older people, more technologically illiterate people can’t do that, and I think deepfakes are kind of the next step in that. But there’s also technofixes out there to address deepfakes and technology that will go frame-by-frame analyzing these deepfake images and you know, the ideas that it can look at things in more detail than a human can tell if they’re fake.
| ZBP | Now, kind of let’s talk about some of the ethical issues that come up, right? Because, you know, when the company starts employing these AI deepfake, you know, detectors or other kind of fraudulent detectors – what kind of are some of the ethical issues we’re seeing the companies encounter when using them?
| II | I think there’s a big concern about the data that’s being used to feed these protective AI algorithms and also you know whether or not these tools might lead to more breaches of private data. I think we already live in a world where there are a lot of breaches of private data, and the concern is that this will only worsen that issue. But I think really the question with that is before we get ahead of ourselves is, you know, will these AI technologies be affective in fighting AI scammers and if they’re 100% affective, I think maybe I’d be less concerned about the privacy component if they’re really good at protecting insured parties and insurance agencies and utilize them well, you know, then I would be less concerned about the privacy.
| ZBP | Do you see or do you think that there’s an issue with how regulators who are obviously outside of the insurance company but still, you know, very involved with how these companies process information – is there a place for the regulators and then the insurers to kind of balance the innovation with the protection of the consumers?
| II | Yeah, you know I think no matter what regulators do, they need to be really cognizant of the fact that whatever they do, these scammers are going to be on the cutting edge of generative AI and they’re going to be using it in inventive ways to accomplish their schemes. And if they’re left to do that without any intervention, you know, people are going to get scammed at the end of the day and you know, they may be left in financial ruin because of that. I think that’s good motivation for regulators to be willing to experiment a little bit more and be willing to take risks that maybe they adopt the policy that doesn’t quite work out but at least they did something. I think its important they get ahead of this and get inventive and rely on experts who I am not and try their best.
| ZBP | Yeah and one of the things I oftentimes close with – as we look forward, asking you to pull out your crystal ball and kind of look, where do you see some of the biggest opportunities, some of the biggest challenges in the next five years as we talk about the use of AI, Artificial Intelligence, specifically as it relates to the filed of insurance fraud.
| II | You know I think one of the biggest challenges will be here for us in the legal industry. You know, the use of deepfakes, falling for deepfake scams, those are presumably going to lead to lawsuits at some point and those are lawsuits on an issue that really hasn’t been heard before in the Court and we’re going to have different stakeholders fighting to get the law to work for them on this issue when it comes to deepfakes. Should the Court be more understanding when an insurance company falls for a deepfake scam, should we hold them to a higher standard and expect them to be the ones to recognize these scams because they’re such a critical issue. Different people are going to want different answers on that and they’re going to be fighting to shape the law, you know, according to their interest and you know, we need to be aware of that, we need to be educated about that and you know, do it in a way that I think works for everyone.
| ZBP | Right. Well Ian, I really appreciate you taking the time to join us today, to talk about Artificial Intelligence and the use of insurance fraud. I appreciate you taking the time and hopefully we’ll have you back someday.
| II | Yeah, thank you. Glad to be here.