Artificial intelligence and insurance: Business risks

Insurance recovery partners Carolyn Rosenberg and Anthony Crawford delve into artificial intelligence, the risks associated with companies using AI in their business practices, and the different insurance products that may help mitigate those risks.

----more----

Transcript:

Intro: Hello and welcome to Insured Success, a podcast brought to you by Reed Smith's insurance recovery lawyers from around the globe. In this podcast series, we explore trends issues and topics of interest affecting commercial policy holders. If you have any questions about the topics discussed in this podcast, please contact our speakers at Insured Success at reedsmith.com. We'll be happy to assist.

Carolyn: Welcome to our inaugural 2024 Insured Success podcast. My name is Carolyn Rosenberg. I'm a partner in the insurance recovery group in our Chicago office and I'll be in conversation today with Anthony Crawford who is also a partner in our insurance recovery group out of New York. Our topic today is AI, artificial intelligence and insurance and you can't help but see word of that, Uh no pun intended, in social media, in a newspaper, Uh certainly the World Economic Forum, recent report indicated that AI powered misinformation is the world's biggest short term threat. There has been a recent 2023 expert study on progress in AI um out of Oxford University that uh created some concern that artificial intelligence could take jobs fifty years earlier than experts predicted. There's been the challenge of AI and the worldwide concerns as well as the promise of AI. The National Security Administration, the NSA, has enlisted AI to fight global hacks. The um there's been stories that AI is helping catch breast cancer earlier because of technology and that AI is being used by nonprofits to help find missing Children. So you've got both the, you know, the horror stories, the concerns the potential exposures. And then of course, on the upside, the promises of AI. Anthony, perhaps you can start us off by indicating, have there actually been any reported lawsuits regarding AI that give us a better handle on how these exposures are being translated into potential causes of action?

Anthony: Yes, Carolyn, thank you. Uh There are, we are starting to see a number of different suits or administrative actions that are coming forth out of uh any of these use of artificial intelligence. One that comes to mind is a recent first with the Equal Employment Opportunity Commission uh having settled a lawsuit uh against an organization called iTutor Group Incorporated. And the basic gist of that was that they were, uh iTutor was using software, AI generated software, that automatically rejected older applicants in violation of the Age Discrimination and Employment Act. And so you can see that this is one of those uses of AI that ultimately cause an organization some liability. And one of the interesting things is that this is this organization had they used pencil and paper or just regular bodies and made the same mistake, they would also face that liability. But the twist here now is that they were using this software that caused that exposure.

Carolyn: Interesting. And have there been any defamation or IP related exposures as well in terms of any potential litigation that you've seen?

Anthony: So on the IP front, yes, there, there have been a number of concerns and you can see this generally in the media about the the use or of AI in this, particularly in the copyright realm and some of the implications there. You're seeing where organizations have to be cautious into making sure that their use of AI isn't violating those rights. And then the uh the other side, you're also seeing where there are complications where artists are actively using AI to help generate works and sort of the implicate the IP implications there as well as you know, other contractual implications where you know, if a particular artist is contracted to do something yet they use AI. Is it the artist doing that? Is the AI doing that? What are the implications there?

Carolyn: Yeah. And I know lawyers themselves are not immune. There have been some well publicized reports about one would call it hallucinations of AI that, that are reporting, you know, uh particular precedents of cases um in court briefs and then upon further examination, um, they don't turn out to be either accurate or in some cases even in existence. And that of course, is, is causing some consternation and concerns about the use of AI in legal research and in briefing. And you could certainly imagine that with respect to any misrepresentative or misinformation, particularly on financial reports that companies and reporting earnings or in reporting on developments could face some potential liability, breach of fiduciary duty. You know, in addition to the kinds of employment, defamation, consumer related IP related exposures that um, Anthony and I have talked about so, you know, but Anthony correct me if I'm wrong, I don't think we've yet seen any coverage, insurance coverage disputes with regard to AI. Is that your view as well?

Anthony: Yeah, Carolyn, I think that that is right. I think that we're seeing a lot of underlying claims that definitely have the potential to be something that would covered by insurance policies. But as of yet, we're not seeing any flood of coverage related litigation with regards to insurance for AI related incidents,

Carolyn: Right, and when we say, ok, there could be these exposures. Now, what do we do about it? And insurance is one potential tool for protection. You know, it, it, it doesn't appear that we need to sort of run out and buy an AI insurance policy. I think some of the ones we've seen on the market have been really more specific for AI companies. But if you're a, an organization or a corporation or a business or a fund that is not AI specific itself but is potentially using AI, what kinds of policies would you traditionally look at to see if there might be coverage for potential AI-related exposure?

Anthony: Carolyn, that's an excellent question. And, you know, I will note, of course, the insurance industry would love for policyholders to run out and buy completely new products. But the reality is uh a as with a lot of the things that we've discussed right now, absent some exclusion, these things should fall in to the traditional policies that already exist right now, such as your areas and emissions policy, your products, liability policies, your employment practices, liability policy. So for example, the EEOC claim that I gave you that at a basic and, and you break it down, it's still a basic age discrimination case, something that is traditionally covered by the employee practices, liability insurance policy. The only difference here is that this company used AI generated technology uh to violate the rights as opposed to using an actual person. You know, some of the other things you may see. For example, if a manufacturer is using an AI Suite to put out a product and there's something wrong with the algorithm and there's ultimately a defect with that product, it doesn't matter that AI was used or not, it's that there's a defect with the products and that will fall traditionally under the products liability type of policy. So the biggest takeaway from this is that entities and policy holders should, you know, not so much if faced with liability, not get so caught up in the AI aspect of it. But you know, the actual facts of the took the alleged violations and then look to their policies that they have in place for any available coverage. 

Carolyn: That's very helpful because the, you know, traditional due diligence that one would do, and of course, we do a lot with our clients representing policyholders of reviewing the policies, making sure there's no exclusion, broadening definitions. And of course, being on the lookout if there are any quote AI enhancements that might come on the market, obviously you always want to keep vigilant in a rapidly developing uh field here, but there's no need for sort of immediate running out and, and looking at a policy if you have the traditional coverage that should um hopefully protect depending on the kinds of exposures um and causes of that may that may arise. But all that being said, we do know that AI is being utilized more in the underwriting in the insurance underwriting phase. And there have been some concerns about that from a state regulatory uh perspective. Anthony, what's your sense of what's going on there?

Anthony: Ok. So in essence, a number of states have recognized the potential pitfalls of using this technology in the underwriting process. And what they have done is to start to impose regulations that will safeguard consumers from sort of automatic pricing schemes used by AI that will take into factors that, you know, some not traditionally used in pricing out insurance policies such as social media feeds or uh you know, this uh a potential policyholders, social presence online. And so what these regulators are looking to do is to make sure that the insurance companies maintain, again, maintain and remain vigilant in their underwriting process. And you know, are accountable for any discriminatory practices that ultimately may come from the use of AI. Uh For example, some of the traditional things that we've seen in the past is using occupations or zip codes or education in pricing out insurance policies uh where states have regulators have pushed back on those practices. You're also seeing a rise in the use of AI in actual claims handling and concerns. There are that when you have this automated process looking at claims, uh there is the potential for the entities to get it wrong and unnecessarily deny claims. So you're seeing there's a couple of uh class actions that are out there against uh a couple of largest health insurers in the country regarding the use of AI in their claims handling in the, in the medical insurance field. And that's something that we'll keep monitoring here as you know, as part of our insurance recovery practice to keep a look out for things like that.

Carolyn: It would be interesting too in the same way that if you were to get into a coverage dispute with an insurer to ask the question whether the insurer and its claim handlers are using AI in reviewing claims and it might be a source of interesting discussion or discovery with respect to where that may lead um as among the tools that advocates may have for their clients in this situation.

Anthony: I agree, Carolyn. I think that you are going to see that question becoming a standard interrogatory or, and uh document requests in civil discovery. Uh That's, that's just, it's going to have to be asked and insurers are going to have to answer that question.

Carolyn: Yeah, it's, it's, it's clearly a a brave new world between the potential exposures that uh insurance coverages, but sort of one potential protection and certainly one to consider and be vigilant about uh because there are always enhancements to consider, there are reviews to be done to just make sure. And in the event that there is a coverage dispute to just be this vigilant on the use of AI in both what's been done on the underwriting end and on the claims end, Anthony, any other dos and don'ts or ways in which to be vigilant as we wrap up this podcast from your perspective.

Anthony: So for policyholders, you know, it as with everything, we always advise, look, insurance comes way way later down the line. All right, what is your, your, you know, the starting point for entities should be sound policies and procedures. So if your organization is going to employ AI, the organization should have a firm understanding of how it's going to be employed, what are the risks that are involved with that employment, and how they are going to mitigate those risks on a practical aspect, before some potential claim comes up. Now, obviously, life happens and should an entity be faced with a claim related to AI, I would give them the advice that I would give them for most any other claim. Ok. Give notice early, get help from some type of insurance professional. Make sure you are organized, make sure you have documentation. Uh make sure you have good lines of communication both within your organization with your broker or whatever advocate that you're using as well as with the insurance company.

Carolyn: Yep and coordination is key. Um Knowing also about attorney-client privilege which you do have with your coverage lawyer you do not necessarily have with, you know, other other professionals. Just be vigilant and monitor. Keep up with us. Uh We will be doing more on AI and insurance in our Insured Success podcast along with other topics. We work collaboratively with our colleagues who are deeply involved in AI guidance monitoring guidelines and all other kinds of good stuff. But we thank you for listening today. Anthony, thank you for being in conversation and we look forward to our next chat with you. No pun intended. Take best care. Thank you.

Outro: Insured Success is a Reed Smith production. Our producer is Ali McCardell. This podcast is available on Spotify Apple Podcasts, Google Podcasts, PodBean, and reedsmith.com. To learn more about Reed Smith's Insurance Recovery Group, please contact insuredsuccess@reedsmith.com.

Disclaimer: This podcast is provided for educational purposes. It does not constitute legal advice and is not intended to establish an attorney-client relationship nor is it intended to suggest or establish standards of care applicable to particular lawyers in any given situation. Prior results do not guarantee a similar outcome. Any views, opinions, or comments made by any external guest speaker are not to be attributed to Reed Smith LLP or its individual lawyers.

All rights reserved.

Transcript is auto-generated.