How Can Insurers Turn AI Adoption into Measurable Value?

Kannan Amaresh:

In the last two years, as the industry has been looking to adopt or pilot AI, they have seen certain gains. Obviously, if you think about insurance industry, it’s potentially, you could say it’s the guardian of the entire global economy. If you think about that, then this industry has got more to do with the risk, and more to do with how do you mitigate and manage the risk. So to that extent, it has kind of moved from pilots to the core enterprise capability. The biggest barrier is, in scaling AI is actually the old legacy system transformation. I think our chairman said it very clearly, it’s the deployment gap. So to that extent, you would notice that the insurance industry has been battled with legacy transformation for a very long time. But this opportunity with AI scaling is going to actually take them forward much faster, in my view.

Jeff Kavanaugh:

AI has permeated insurance faster than anything we’ve seen before. But here’s the challenge. For many carriers, this adoption still hasn’t converted to broad-based measurable value. Volume two of the Infosys Global Insurance Journal is aptly named Unlock AI Value. In this interview, we examine what creates value, the barriers, and what insurers can do to close the gap. I’m Jeff Kavanaugh with the Infosys Knowledge Institute. I’m joined today by Kannan Amaresh, Global Insurance Leader for Infosys. Welcome, Kannan.

Kannan Amaresh:

Thanks, Jeff. Happy to be here.

Jeff Kavanaugh:

Before we get to AI value creation, let’s establish where the industry stands today. How would you describe the current state of AI adoption and insurance, and does it match the media narrative?

Kannan Amaresh:

We looked at the industry pre-ChatGPT and post-ChatGPT in my view. So if you look at the starting point as being like November ‘23, and if you think about that, insurance was always seen as a catch-up player industry to many other industries, but post November ‘23, I’ve been pretty surprised to see the rapid adoption of AI. It’s no longer just pilots. It’s about how do you get more value out of the scaling the AI across the value chain. To that extent, I’m actually happy to see insurance industry putting AI as part of their core enterprise capability rather than just on the fringes.

In the last two years, as the industry has been looking to adopt or pilot AI, they have seen certain gains. Obviously, if you think about insurance industry, it’s potentially, you could say it’s the guardian of the entire global economy, right? So if you think about that, then this industry has got more to do with the risk, then more to do with how do you mitigate and manage the risk. So to that extent, it has kind of moved from pilots to the core enterprise capability. The biggest barrier is, in scaling AI is actually the old legacy system transformation. I think our chairman said it very clearly, it’s the deployment gap. But this opportunity with AI scaling is going to actually take them forward much faster, in my view.

If you look at the competitive advantage for insurance industry, it’s all about measuring the risk, probably preventing and mitigating the risk. That’s the core of underwriting. And then if you look at the claims side, if you look at how do we actually get claims to be done efficiently. Then that gives you an opportunity to use AI to really provide digital view for the claimants so that they understand what’s going on when they submit a claim. So I’m excited to see how this scaling of AI is the center of all conversation.

And recently Travelers publicly talked about they produced something called TravelersLLM which is to harness their own data and create a LLM for their own enterprise to make sure that scalability is there for AI. I actually call it the AI glue.

Jeff Kavanaugh:

They’re excited, insurers, carriers are jumping in, but they seem to be stuck between these promising pilots and payoff at the enterprise level. And that gap seems to have become a defining story. What will ultimately drive the value and what does it tell us about where AI and insurance is headed?

Kannan Amaresh:

You probably would have seen the Evident AI Index recently, the 2026 was launched. And when you look at the top thirty insurance player, they’ve given you a very nice view of where is the scalable opportunity. That goes back to the talent plus token economics in my view. Or rather, if you want to put it simply, the human plus AI. How do you amplify the human capital? Once that part is understood, the second part is really around legacy transformation, data for AI, getting the data governance right, making sure the enterprise data is available across the value chain. The key is to really unlock the data for AI and then, of course, put human plus AI in the front of the transformation.

Jeff Kavanaugh:

Insurance has always been about risk aversion and assurance, judgment is also very important, obviously, built over years by underwriters, claims professionals, and actuaries. How should insurers deploy AI to preserve that valuable expertise, yet still accelerate the next generation of talent behind it?

Kannan Amaresh:

When you think about insurance business, it’s all about risk, underwriting the risk. And when you talk about risk, it’s the economic value of risk. If you take P&C property and casualty as an example, focus is obviously AI needs data center. So there’s going to be a lot of rush on building data centers, but then guess what? Even the data centers need insurance. So whether it is flood or whether it is any act of God or anything that goes wrong with the data center needs to be protected and insurance has to be taken. But then you look at the economic value of that risk. So that’s where I think the focus is on.

You’re right in the focus that the industry has always had the data and the expertise. But today every insurer is looking to unlock that expertise via AI so that the repeatability is faster, part one. Part two is about scaling so that the enterprise, most of the insurers, if you look at it, are global, right? So if you take the top 50 insurers in the world, they operate probably about 150 countries on an average. So what that means is that they have to scale their expertise faster. For that, you need the data set to be available across the enterprise.

Then comes this whole new risk velocity of new data that is coming. For example, as I said, if today everybody is putting up a data center, there’s lots of new data that’s coming up. Plus the way you look at the geospatial data, the amount of data today is available for an insurer has gone up by probably, I’m guessing, 25x, go back like 20 years back. So which means that their ability to intake the data, obviously get value out of the data using the human expertise is all combined together, is the recipe for scaling AI, in my view.

Jeff Kavanaugh:

You’d mentioned before how it’s exciting to see AI in the core. So it’s embedded within claims, underwriting, pricing, and customer interactions. This also, though, raises the stakes on regulation, trust, and accountability. How should insurers think about that responsibility as adoption grows?

Kannan Amaresh:

Generally think about insurance business, it runs on trust. The fundamental principle of insurance business is trust. It gets even more amplified with focus on AI because you have to make sure that there are no biases creeping in. You have to have an explainability to not only the regulator but also to all the stakeholders. So in that sense the cornerstone of the industry is trust, but how do you amplify with AI? So that comes with explainability.

In fact, the way I think about it, there’s this huge focus on fraud insurance. Deepfake or like impersonation and so on and so forth. But how do you actually use AI to actually prevent that and do a great job in making sure that you have a better trust with your customer, especially older generation when they come for long term care or claims.

Jeff Kavanaugh:

Insurance executives seem to have many AI opportunities to cover. If an insurer asked you where to make a serious AI investment, what and where would you tell them to focus?

Kannan Amaresh:

I think the biggest focus, generally speaking, they want to get the underwriting right, but that’s the hard problem to solve immediately. I would rather say focus on the claim side and the customer experience side. Start with customer service, customer experience. If you think about customer service, it’s all about speed. And where your customer side is taken care of first. So focus on AI deployment, helping your customer service side, make sure the experience is seamless, provide them with a lot more information, provide them with all the answers they’re seeking.

When COVID hit, lots of customers started calling these customer service agents to know whether they are covered, not covered. You got to enable that much faster. So customer service first, claims second, maybe look at all other internal operations, how do you make everybody access data, everybody has ability to collaborate, so all of that internal process next, and then maybe look at underwriting as a hard problem to solve.

Jeff Kavanaugh:

AI access, while it’s a differentiator now, it will not be so for long since the technology is becoming very available. What will determine whether an insurer converts this growing AI access into competitive advantage?

Kannan Amaresh:

If you look at the competitive advantage of the industry is to minimize losses, generally speaking, right? Like they don’t want to be above the combined ratio of hundred if you generally talk about the industry to start with, which means here is the provocative question I was talking to some of the clients saying, okay, if you have matured AI to the highest level, would you combined ratio be like 40? Instead of 85 or 80, generally speaking, because that’s a metric which the industry tracks. What would that mean for your pricing strategy, your products, and so on and so forth? So the little bit of provocativeness on that side.

But coming back to the industry will be well served if they share more and more of the risk, more and more of the data, so that there is a little bit of more of the industry view first, Jeff, and then comes to the competitive advantage within the industry. There I would count for how do you make sure that the claims side is extremely smooth. No matter whether it’s individual or corporate or whichever SME, customer, whichever be it, claims will be the first place I would focus on saying, can you build differentiation there? Then comes the underwriting piece in terms of market, access, products, and so on and so forth.

Jeff Kavanaugh:

AI investment is growing like crazy, especially on the back end. But as it grows in the enterprise, the harder question becomes measurement, measurement of value. How should insurance leaders distinguish between activity that looks promising and the transformation that delivers that measurable value?

Kannan Amaresh:

In fact, I was just talking to a couple of the CXO couple of weeks back. In fact, one of the CIO called me and said, hey, you guys as Infosys, you guys are adopting AI across all vertical, you guys have done so many things within the firm. How do I look at my talent and token economics? Because I’ve been in the business twenty-seven years, Jeff. When I look back the business case, maybe like ten years back, people used to do a business case which is very specific to like, okay, this is what we are going to save, this is what we’re going to make revenue. But today, the AI scale adoption is going to give you much more progressive payoffs than what you see right now.

So in that sense, what I suggest to our CXO community is to look at business case at 5-to-7-year view. Don’t look at just 3-year view. Don’t look at just three-year view. If you have got a foundational data AI done, you can easily launch products much faster. One of the P&C clients is looking to launch, let’s say, a new product. It takes them 15 months to launch a new product. 15 months. So the incremental value the enterprise would get in investing in AI is not just the business case for today, but how do you actually prepare it for 3-to-5,7 year view?

And to that extent, I would say again the fundamental is get your customer service right, experience right, so you can keep scaling. And likewise you get your internal operational efficiency right with the talent and token economics. So that gives you a flexibility, of course, working with partners. And then comes the core of underwriting so that your combined ratio, when you project it back, it’s not looking that bad.

Jeff Kavanaugh:

AI is clearly disruptive, not just for insurers, but for the risks they cover and the partners who serve them. Given that disruption, why are you optimistic about what AI will provide for the industry’s future?

Kannan Amaresh:

If you think about insurance industry, it’s all about managing disruption and managing risk and managing uncertainty, which is why I coined the word this whole guardian of the global economy, right? So if you take that fundamental view that insurance industry is prepared to handle all kinds of disruption, this actually you know, as I said, cyber risk. If you look at cyber insecurity, as we were debating in the Global Risk Report board, the cyber insecurity is actually providing more opportunity for cyber insurance, right? If you think about anything volatile like climate change, again more opportunity for insurance business, both preventing, protecting as well as potentially working with the industry to mitigate as well. So every risk is actually an opportunity with insurance industry. And of course AI also as I said, AI actually both ways is an opportunity for the industry.

So if you look at that perspective, I’m extremely optimistic because there’s going to be more new products that’s going to be coming out of the insurance industry to support the global economy. Another example I’ll tell you is flood insurance in the US. So crop insurance is another big example where we are saying, okay, digital farming, how do we actually predict with data? How do we actually create more opportunities? This is another reason why I’m so excited because we seem to be in the middle of this whole AI adoption as well as managing risk for the global industry. So you can’t ask for any better industry, I guess.

Jeff Kavanaugh:

Well Kannan, thanks for the conversation and that energetic uplift at the end. I agree with you and I believe that insurance quite often is overlooked in the broader society, even broader business, not only its importance, but the excitement because it lays the foundation for people to take risks. Because you’ve got somebody who’s the guardian.

Well, all great points. And to go deeper on these points, check out the Insurance Journal, available now.

Thanks for joining us and until next time, keep learning and keep sharing.