Latest / Leader's Edge / Cyber Exposure in the Middle Market
Transcript
- 0:00Speaker 1: And I think that's where we're gonna have to start pushing for controls like an MDR or EDR because AI is gonna be sophisticated enough if leveraged by the threat actor. They're gonna be able to then circumvent or bypass or have the ability to bypass that particular control.
- 0:21Speaker 2: Welcome to the Leader's Edge podcast. I'm Sandy Laycocks, editor in chief of Leader's Edge. In this cyber risk conversation sponsored by Nationwide, I talk with John Butler, cyber product lead at E Risk Services. We are focused specifically on cyber risk to the middle market in this conversation, And we start off by discussing the cyber threat landscape for this group, including the growth of third party risk.
- 0:49Speaker 1: Right now, the market is seeing evolving threat complexities. That is due to the rise in AI being leveraged by threat actors as well as geopolitical instability and supply chain vulnerabilities that we're now seeing can bypass traditional defenses for for those that have proper or what would be considered best practice security controls.
- 1:22Speaker 2: Can you tell me a little bit more about the supply chain vulnerability?
- 1:25Speaker 1: No. Absolutely. So what we've been seeing over the last, say, eighteen months, or two years even, is that we have different components of supply chain. So it's not it's not supply chain in the in how one thinks about distribution, of product and so forth. Supply chain here is has to do with if an insured has vendors that they are doing business with that have that are have some sort of connectivity between the insured and the vendor or just or even having vendors that that provide some sort of nontechnical service, they may end up being vulnerable because of they could be vulnerable based off the threat actor utilizing the vendor as as impersonating them to then commit some sort of fraud, and that and that goes to the social engineering as well as wire transfer fraud that happens in today's world.
- 2:36And that that that goes to the supply chain as well as your software. Sometimes you're gonna see software as a service as being a prominent way of of of our world as as as a corporate commercial operating for our for our business to operate. And so you you're starting to see ripple effects of threat actors that are able to get access to the insured through the vulnerabilities of the of so called software software application that's being provided by a third party or a vendor of theirs.
- 3:20Speaker 2: Do you feel like the the middle market segment is adequately covered for this?
- 3:27Speaker 1: That's a great question. And that's going to be one of those questions that I do get a lot, especially being at especially talking with brokers and other other professionals within the insurance industry that look at that that sees cyber as a very young, even though it's a young product, even though it's been around for almost you know, little over twenty five years, we're just now starting to see some maturity within the claims claims data. Mhmm. But it's been around for for quite some time. You do you do see that because they are middle mark because middle market is really makes up a lot of your small business, and even maybe your more more national with some international presence, you may have you may have that, you know, limit adequacy question, where do they carry enough limits that reflect, you know, their budget tolerance or or even modeling their exposure or have the ability to model their exposure, that tends to that tends to form some gaps.
- 4:41And so it's very important for, companies in the middle market to really understand their digital assets. Digital assets is very key, because those are the access points, that I that I would that I can talk about when we're talking about the security controls. So if so there's have been there a lot of times, and you see this when your claim comes through, is that an insured or someone that is, you know, middle market middle market business, they may have certain controls in place, practices, which would be like MFA, which is multifactor authentication or endpoint detection response. Or even next step further, they may have a, managed service provider, providing a managed detection response kind of platform, which allows for, them to monitor the activity on their infrastructure so that they can then be alerted and notified. And then if there is a vulnerability, can be once it's been detected, can be then remediated so that there isn't a issue with a threat actor exploiting that vulnerability.
- 6:00And sometimes what happens is that they may not have the, you know, the techno you know, the the tech personnel to be able to then service those sorts of those sorts of risks or the or the exposure that comes from all, you know, the just the infrastructure of their of their network. And so that's where we're seeing, some of, some of the gaps, in not just in coverage, but just gaps in the the the resilience or the cyber resilience of their of their, of their defenses.
- 6:40Speaker 2: We then turn to AI, discussing what exposures the middle market has with AI use and where it might show up unexpectedly in their claims.
- 6:52Speaker 1: Artificial intelligence is going to be, I think, the next the next emerging risk that cyber was, say, three, four years ago when we were talking about silent cyber.
- 7:06Speaker 2: Right.
- 7:07Speaker 1: We're probably gonna end up talking about silent AI exposure, which that means that not only do we feel that the cyber policy is right now in its current state that's best suited, but you are gonna start seeing because of the use of AI, that's gonna broaden out where other lines of coverage is gonna get you know, will come into play, especially if you're talking about AI and the use of emerging for there to be maybe, say, bodily injury. So when you talk about autonomous technology where you have, you know, unmanned driving cars and somehow there's a vulnerability within the technology that allows it to be autonomous, now you got now you got some risk there where are you gonna have an e and o claim because of the bodily injury? Is it gonna fall under the general liability, auto liability? You have also then if there's property damage, property coverage, or is it going to be, you know, other? And then also then you have the situation of the Ds and Os, you know, the d's and o's of those companies.
- 8:22Is that going to have an emergence of an effect on the directors and officers' coverage as well as EPL using AI with respects to employment platforms or software that where they use AI. And so now you have that bias discriminatory exposure that could come into play.
- 8:45Speaker 2: So do you feel like that these all of these risks will creep into a cyber policy per se? Or that is what you're thinking might happen?
- 8:56Speaker 1: Yeah. Right now, the cyber policy right now may not provide affirmative AI language. However, the policy would respond to a cyber event if AI is if AI is Right. Being used by the threat actor or if AI is is being used by the actual company. The the issue with the from the usage of AI, that springs the privacy exposure.
- 9:29Because of the way that AI generates their outputs, you need to provide it data. And then that's how it learns. Well, what kind of data are they inputting? And if what and based on what data, what are the governance protocols are that the insured policyholder has in place that are going to be mandated by state statutes. And that's one thing that we're starting to see starting in 2026 next year is that there are already gonna be some states that are going to be online for regulating against the governance of of AI Mhmm.
- 10:10For that purpose. And that that just kinda gives a little bit more an enforcement mechanism just like we've been seeing in the prior years with BIPA, which is the Biometric Privacy Information Act, where we've been seeing a lot of class action lawsuits that come from that particular regulation or or statute. Mhmm.
- 10:33Speaker 2: Yeah. It does seem like and I know that the the class action lawsuits are there have been some based on the data privacy laws recently. So that I yeah. It seems like it's gonna be a little bit of a messy a messy
- 10:51Speaker 1: It is. It is. And and going back to the landscape of the market, it's gonna be interesting in how the models start changing when we start incorporating more longer tail claims into those models. I think right now, a lot of the models, there's assumptions within those models where where they they are contemplating, but I think it's not contemplating enough of that particular part of of the of the risk. And privacy for me is like the elephant in the room just like e tech e and o is when we add tech e and o to our cyber policy in the market, in our industry.
- 11:37That to me is the big elephants because they provide longer tail than what we have been seeing. With ransomware, it that's technically first party, a lot of the expenses you're you're dealing with. And so that, usually, those have shorter tails, a lot of the first party coverage that we have been seeing with ransomware. But it's it's that liability piece, whether it's the privacy that comes from the liability or or now that what we're what we're gonna be dealing with is more regulatory, I do think is gonna come online to then perpetuate that privacy, exposure for for the long term and then providing there to be more longer tail evidence when it comes to when we're looking at the models for for cyber.
- 12:33Speaker 2: Next, we look at AI from the threat actor side and how advanced threats utilizing AI are being taken into account by underwriters.
- 12:43Speaker 1: We are starting to see a lot of the AI being used by threat actors. They're leveraging it to basically provide, more hyper realistic. And I use hyper because you'd be amazed, like, you'll get emails or the emails that are drafted are almost exactly in the tone
- 13:06Speaker 2: Yeah.
- 13:07Speaker 1: Of who that person it may come from. And so it can be it's gonna eventually get very, very, very hard to distinguish, the phishing emails as well as any sort of spoofing or deepfake ploy Yeah. To be able to get you to do something for them. And in this case, most likely, it'd be the transaction of money is where the theft is gonna be coming from. Also, on the privacy side, we probably are going to start seeing issues where AI is being misused by employees, where where when you misuse it outside of, say, the the the policies and the governance of the employer, you put yourself at risk for any sort of leakage or any sort of data privacy violations when it comes to the idea of the authorized use or or unauthorized use of of that particular data, that's being used for purposes of of output, when it comes to AI.
- 14:20And the other thing is also that we are we're really seeing is the ability to, enhance, or, you know, really truly engage in in in best practices on hygiene and the and the controls around all of your data as well as your, systems. Being able to, put in place proper, privileged access, meaning oh, maybe restrict access first before giving everyone access. You kinda go the reverse. Every you know, go to more of a closer to a zero trust, privileged access approach. But as well but also making sure that you're having MFA deployed across your enterprise, making sure that you have, some sort of endpoint detection, but that's deployed across your whole entire enterprise.
- 15:18So you're monitoring every single digital asset that may have some sort of interaction with the, with the connectivity of the Internet so that you're managing and monitoring that that traffic. Those probably two things are gonna be very, very important as we learn more about AI since AI is completely accelerating the the risk models within the cyber within the cyberspace. Yeah. And it it it's it makes it for an interesting day. Definitely, it's not dull.
- 15:55Not dull? Not dull.
- 15:57Speaker 2: Given all of that and given all of that risk and how it's accelerating and how, obviously, the underwriting has to keep changing to to incorporate all of this in, do you feel like the market is is adequate? Is it is our premiums adequate to to provide the right amount of coverage for these companies that are actively, you know, utilizing these tools?
- 16:21Speaker 1: I'm gonna have to speak speak it as if it's it's my own professional opinion. We're definitely in a soft market, and we're in a soft market because of the hard market we went through because of the ransomware. And and the 2020, we just saw that's when we really truly saw spiking claims and frequency from severe you know, from the idea of severity and frequency together, we saw that spike. The industry felt it, and we needed them to basically, respond, and we did. We increased rates as as much as double digit plus.
- 17:01And so so now we're in a market where we may not be at that stage where we were pricing wise pre 2019, but it feels like it because of the softness and the competitiveness of the market with new entrants and also the the ability now for increased capacity. Mhmm. Before two, three years ago, capacity was an issue because of the because of the the law the number of losses in that eighteen eighteen month to twenty four month period. But I do feel that we are we are maturing enough to understand, that we have now tools to be able to utilize and leverage, with respects to the underwriting. And, also, we've been getting really, really good at the, at the application process and what to look for and really laser in on, real specific controls that matter.
- 18:05Because if you don't, then you are gonna see the claim. Example, during the the years that we're seeing heavily ransomware, we've been it's the the not just the cause, but the reason for it was the the ability to access the insured's network. And so MFA was critical to be able to be like your first line defense, to be able to to be able to at least give yourself a good fighting chance against the threat. Now with AI, that makes it a little we have to go next step further. And I think that's where we're gonna have to start pushing for controls like an MDR or EDR approach because AI is is gonna be sophisticated enough if leveraged by the threat actor.
- 19:00They're gonna be able to then circumvent or bypass or have the ability to bypass that particular control. Mhmm. And then for us, we basically can we can on our side, we we have tools ourselves to be able to understand from an outside in view of the risk, meaning the the, the perimeter that we can see the outside is kinda like going window shopping. We can at least get a small small look at look at to see, like, is their outside perimeter defenses look like? Mhmm.
- 19:39And that helps us use that data in conjunction with the application as a way to complement our underwriting. And that's being that's being done, throughout the marketplace. We're we're just we're not the only ones, but that's something that the market has seen as being a way to help with complementing the underwriting because it is gonna get really, really sophisticated as time goes on.
- 20:10Speaker 2: Finally, John shares some key advice for brokers looking to secure coverage for their middle market clients.
- 20:17Speaker 1: Biggest takeaways, I would definitely hone in on or focus on the services that any carrier is able to provide their insured. If you're a retailer or even if you are a a broker, that to me, I think, is going to be the value add to your insured is what is what is the market providing in terms of services that can help on a pre free breach basis, but then also understanding, like, what are what are the preapproved post breach, and aligning some of those some of those elements, so that your, so that the insured is better prepared, especially when it comes to, an an instant response plan. It's good to have in place, you know, who who are we calling first, and who do we have to engage with first. A lot of the markets, in this space have hotlines where they can call, to get ahold of a breach counsel, and then they'll quarterback with the insured. And then also with the carrier, they'll quarterback what needs to happen.
- 21:31And so that a lot of that stress could get be relieved from the insured knowing experts that are now can take take that on for them and help them through that that incident because it can be very stressful in any with any incident in that respects. But that would be the main thing that I would truly focus on as well as as, of course, coverage, making sure that that you're getting the most coverage across the board with respects to third party and first party, for the client, because there are gonna be situations where, you're you want them to have the more comprehensive coverage versus, what we do see is, like, sometimes you could get coverage for maybe a lesser premium, but it's like a it's an endorsement that gets put on to maybe an existing policy. I would not I I would be someone in my professional's opinion, say that would be a way that that wouldn't be what I would try to educate the insured with. I would try to educate them with a more comprehensive, approach with the with the with the with the coverage, but then also with those, services, that we that, provides in some fashion or form.
- 23:03Speaker 2: That was John Butler, cyber product lead for Nationwide's E Risk Services. Thanks for listening. I hope you enjoyed our conversation. Subscribe to the Leader's Edge podcast at leader'sedge.com.