TV

Paul Lucas  00:00:01
Hello everyone, and welcome to an AI focused special edition of Insurance Business TV brought to you in association with Tokyo Marine HCC Cyber and Professional Lines Group. Now, while AI improves cyber security defenses, it also introduces new risks. Indeed, emerging AI technologies are likely to further shape risk management and insurance strategies, and these will be among the topics discussed by today's panel. So let's meet them. Shall we? We have Eugene Eychis. He is Director Cyber Underwriting Tokyo marine HCC, Cyber and Professional Lines Group and Dan Crane, Senior Professional Broker at RT specialty, so welcome both and to get us started, tell us in what ways AI is currently enhancing risk assessment and underwriting processes within cyber insurance, Eugene. I'll start with you. 
 

Eugene Eychis  00:00:54
Yeah, thanks thanks for thanks for having us. So we started using AI last year, and within our organization, some of these AI tools, and with the latest AI tools, we're able to better understand individual risks. And, you know, during the underwriting process, I mean, we see 1000s of submissions every year. And so what it's able to help us do on the underwriting side is get concise summaries of an applicant's operations, so there's significant time saving there. We're also able to interact with it and ask follow up questions about specific exposures that we're concerned about. For example, we can ask, you know, has there been any past litigation? We can ask, Has there been maybe, any type of new legislation that will affect a particular industry or a specific insured So ultimately, it's, it's helped us get information a lot quicker. It's helped us make underwriting decisions a lot faster, and it's also really reduced a lot of the friction, a lot of back and forth with asking additional underwriting questions. And so I think that's helped save time for everybody involved.
 

Paul Lucas  00:02:18
Okay, so that's the insurer side of the equation. Let's get the brokering side down.
 

Dan Crane  00:02:23
Yeah, thanks Paul, thanks for having us today. I think when you look at the broker side, it's how can we do our job more efficiently, more effectively, in order to bigger, deliver the best results to our policy holders and retail agent partners? And the use of AI can allow the streamlining of policy comparisons, while it's something that we're ultimately going to review and verify all of the information, it's another tool that we can put to use in order to streamline that process, or generating submission emails and further evaluating risks before it gets to the underwriting process. So similar to Eugene, we're able to utilize the same process on the front end to ensure that the carriers are getting the best submission possible, that everything is being contemplated on the front end, and that we can deliver the best solution to our clients.
 

Paul Lucas  00:03:15
Well, looking at some positives with AI there, but Dan, if I can stick with you for a minute, tell me how the integration of AI has actually transformed the cyber threat landscape as well in recent years, and indeed, what implications does this have, both for underwriting and for policy holders?
 

Dan Crane  00:03:32
Certainly. So I mean, I think the biggest hot topic right now will be the increase in social engineering. So within a cyber policy you'll have your traditional first party coverages that will respond to your business downtime, as well as a variety of other components on the third party side as well. But within that, you have your crime type coverages and the social engineering attacks we've seen a vast increase, and really, AI has enabled these bad actors to become more sophisticated. So the messaging that the individual at the company is receiving the email is just much more personalized. It's sophisticated. They're able to utilize AI in order to generate these messages in a attempt to get better click through in order to then compromise these systems. So that's definitely the biggest hot topic, I would say that we're seeing on that front at the moment. 
 

Paul Lucas  00:04:36
That's such a great point as well. Isn't it? Eugene these, these phishing attacks, so much more difficult to spot?
 

Eugene Eychis  00:04:42
Yeah, absolutely. I mean, we see obviously the same thing on our end. And as far as social engineering and phishing attacks, I'll say, as far as kind of the second part of the question, as far as implications for underwriting and AI. I what, what AI allows us to do is analyze vast data sets of information in real time, very quickly, and also offer more precise risk profiles on our individual insureds. And that's been it's been very helpful for us. AI is also on the operating side. It's also helping us address some of the systemic risk challenges that we face. Because these IT systems are so interconnected, it means that a single vulnerability can impact 1000s of our policy holders all at once, and so insurers like ourselves, we address this by by analyzing sort of the metadata that we have and updating our models, right updating our models, and then Taking action based on on on those results. And then lastly, for policy holders, what we've seen is really better outcomes and reduced losses. So when our policy holders work with our Cyber Threat Intelligence Team to address certain vulnerabilities, we've seen that that helps ensure adult when you get ultimately get better coverage and sometimes reduce premiums. 
 

Paul Lucas  00:06:30
Well, clearly Eugene, there are plenty of risks associated with AI can be used maliciously in cyber attacks. Of course, deep fakes and automated hacking spring to mind. So What strategies can be employed, employed to take this on?
 

Eugene Eychis  00:06:46
Yeah, I think there's, there's a multi, sort of multi faceted approach here, and that includes the technology component, but also a human, human awareness component to it. So from a from a technology component, something that we do, and other insurers do, is provide real time continuous it monitoring of our insured systems, and I think we're also very good at staying up to date through our Cyber Threat Intelligence team to make sure we understand what the latest attack vectors are in any type of AI powered attacks, and then just really staying up to date and adopting new technologies such as deep fake detection tools. And then, I think the other main, you know, the other important part of all this is the human awareness factor, which is, you know, I think there's a need to educate users of these risks of AI, such as deep fakes, such as phishing, and just promote, sort of the awareness of any type of suspicious activity. And so one of the ways that we do this, for example, is we provide our all of our insureds with access to security awareness and phishing simulating, phishing simulation training resources.
 

Paul Lucas  00:08:15
And I guess that brokers too need to stay up to date Dan? 
 

Dan Crane  00:08:19
Certainly, yeah, Eugene makes a great point in pointing out one of the free resources, the phishing training that a lot of cyber policies will come with, so that is a big piece to address the employee component, to ensure that as the threat actors and the new ways to attack These employees and dupe them are out there that the employees are educated on it so they can be on the lookout for what is the new way to kind of compromise their systems and gain access to these computer systems. And I think, yes, I think it's really kind of a that would be one component, I think the other component, Eugene also touched on with the technology side, as there's vulnerabilities within these AI models and these large language models that companies may now start to use, understanding and being on the lookout for any threats that can create a bigger risk within their organization. And then the third piece, which we didn't really touch on yet, and it's kind of ties back to the last question on kind of emerging threats is a regulatory component. I mean, I think as you look at the data that these employees and employers are collecting and where is it being stored, what, what is ultimately happening with it, if there is an incident, um, what responsibilities? What new regulations are going to continue to evolve, as we've seen in the past. So that's something to definitely be on the lookout for, on the regulatory and legal side, that organizations can kind of better prepare themselves for by staying up to date on any new legislation that's out there. 
 

Paul Lucas  00:09:57
Very conscious, and I'm being quite negative so far, making you. All talk about the threats of AI, so let's get excited about AI as well, and tell us about the emerging technologies and cyber security that you're most enthusiastic about, Eugene, and indeed, tell me how you see them evolving over the next decade.
 

Eugene Eychis  00:10:14
Yeah, sure. I think when we talk about AI, I think it's important also to differentiate between kind of traditional based automation and real, AI driven capabilities. So most security solutions have actually always kind of used automation or some type of machine learning. It's been around for a long time. I think as far as new or emerging things and technologies. I'll give you two that I think we're pretty excited about. So for example, voice and video fraud detection, right? So it's software that detects if it's a, you know, AI powered voice or video, and then AI powered email security tools. So these tools are going to go a little bit more in depth. They understand the content of the email, for example, they understand the intent and that. They also look at behavior, right? They understand how the user has acted in the past, and how that's kind of different going forward. And hopefully spot those spot those differences, it's potentially stop those phishing attacks from happening, 
 

Paul Lucas  00:11:32
I guess, from the the broker standpoint, Dan, it's just all about having better tools to do your job a little bit easier.
 

Dan Crane  00:11:39
Yeah, certainly. So I think really excited when it comes to as AI evolves and we're able to better you utilize it to compare policies and compare coverages and specific wordings. So that's something that's quite manual and time consuming now, and having the ability to utilize that as a another tool and another resource just to work more efficiently is something certainly exciting, and I think similar to Eugene of evaluating exposures, evaluating risk, and as we're looking at benchmarking and what limits should be purchased, being able to utilize AI in that component of it to ensure that there's adequate coverage in place and that we're recommending the right limits for the right client.
 

Paul Lucas  00:12:27
One more question for you, Dan, if you don't mind just for the other brokers out there, do you have a thought or a tip about addressing AI with their clients? And if so, what would that tip be?
 

Dan Crane  00:12:37
Yeah, I think it would just be ensuring that you understand the entire exposure, and if them utilizing AI within their own organization, or if a third party they rely on utilizes AI, if that increases their exposure, and if that creates a more systematic concern with an impact, a vulnerability to An AI system, or that going down, then leading to you your business being unable to operate as it normally would so, the business interruption side of your cyber policy, as well as the dependent business interruption side for those third parties that you rely on and ensuring that there's adequate and appropriate coverage there. And the other piece would just be overall limits management, with some of the coverages we discussed on the fraudulent instruction, cyber crime, social engineering, those types of coverages that the right limits are there, the right education is there for your users, and just really embracing the resources out there, but also ensuring that all of that exposure is contemplated.
 

Paul Lucas  00:13:41
And Eugene, let's get a tip or a takeaway from you as well, because this is just a massive change for the world, isn't it?
 

Eugene Eychis  00:13:48
Yeah, absolutely. I think this is huge. And I think the way I look at it is is, oftentimes what I see is people are reluctant to change and reluctant to adopt new technology. And I think my kind of main message and takeaway is that people should, should embrace this change. And I think, you know, the world is moving kind of in this AI driven world, and I've seen, you know, I think in the future, successful organizations, successful individuals will adopt this technology, and they'll have good outcomes because of it. 
 

Paul Lucas  00:14:26
Fantastic thoughts from you both. No doubt AI will continue to dominate the headlines. So if you want the latest news to keep up with the latest tech, keep it right here. And Insurance Business, TV.