From Insurance Payer to Prevention Partner

Pete Miller [00:33]: Welcome to the podcast! I’m your host, Pete Miller, and today we have a special episode to speak with the winner of the International Insurance Society’s 2025 Predict & Prevent award. I’m pleased to welcome Pierre du Rostu, CEO of the AXA Digital Commercial Platform.

Pierre leads an innovative team that’s transforming how businesses approach risk – combining cutting-edge technologies like AI, IoT sensors, and real-time satellite data with AXA’s deep risk management expertise. In this episode, we’ll explore:

• How AXA’s platform is helping organizations predict and prevent losses before they happen…
• Why Pierre believes we’re living in an age of “polycrisis” where multiple risks cascade and amplify each other…
• And how the insurance industry is fundamentally shifting from simply paying claims to becoming a true risk prevention partner.

The award will be presented at the IIS Global Insurance Forum in Switzerland later this month, but before then let’s learn more about the AXA Digital Commercial Platform….

Pete Miller [01:50]: Pierre, please provide us with a high-level view of AXA’s Digital Commercial Platform and how a customer could use it to predict and prevent potential losses.

Pierre du Rostu [2:00]: So DCP is a one stop shop risk management solution, allowing businesses to predict and prevent their risks in one place. It combines historic data sets, real time data from satellites or from IoT sensors, third-party software and AXA in-house expertise to give businesses comprehensive understanding of the site and protection. The platform lets customers gauge and fix their vulnerabilities, see threats early and act before damages occur. It helps organizations to stay resilient, keep operations running and most importantly, avoid costly losses.

Pete Miller [02:40]: So that’s very interesting. So you mentioned a number of different kind of technologies. So can you tell us what are some of the technologies being used by the platform to mitigate risk?

Pierre du Rostu [02:51]: Yeah, absolutely. So DCP is built on Amazon Web Services, we have a strategic partnership with Amazon and we use their services quite a lot to build the platform to handle vast amounts of data quickly and with high level of reliability. But on top of that, we use a lot of real time geospatial data from diverse data providers, public and private fleets, to track wildfires, flood risks, climate change, etc. We use a lot of IoTs, so sensors, be it in ships, in planes, in trucks, cars and so on and so on. And of course, quite recently, we’ve been accelerating on generative AI.

For us, it’s a game changer, and we are very much focused on investing on Gen AI because we think it’s going to be an evolution for us.

Pete Miller [03:40]: Yeah, very interesting. The Gen AI angle, of course, is very timely. So your award application talks about the world being in a “polycrisis.” So, what does that mean? Can you give us an example of how these different risks crash into each other in real life?

Pierre du Rostu [03:56]: Absolutely, and you know, in the past we were used to having one crisis after the other, but what we have observed for the past five-10 years is that now it’s not one after the other, but we have crises that are interlinked.

So the term polycrisis describes a situation where many risks overlap, interlink and amplify each other. So, if you take a practical example, climate change creates social instability. Social instability could lead to geopolitical crisis, geopolitical crisis could lead to cyberattack, to supply chain disruption, etc. So, there is no more single event isolated, but it’s kind of a cascade of risks across borders and across industries.

And this is why we thought of the platform model to be able to address the different risks at the same time and make sure we can model and create connection in between climate change, geopolitical risks, supply chain disruption, and so on and so on.

Pete Miller [05:00]: So traditionally, most insurers pay claims after something bad happens, right? So, it sounds like you’re trying to prevent bad things from happening in the first place. So, kind of from a business model point of view, how does that change AXA’s business?

Pierre du Rostu [05:14]: We live in a world where insurance can’t be any more the unique solution to risks. If you look at the numbers: you know, last year, I think it was two-thirds of the losses related to climate change that were not insured. On cyber risk, I think it’s more than 90% of the losses that are not insured. And on geopolitical risk, we think it’s even higher than 90%.

So, it puts us in a corner where we have traditional way of looking at risk and doing big insurance business, but this can’t be the solution anymore. And we, as an industry, have to completely reinvent our model because this is not the model of tomorrow. It’s part of the answer for tomorrow, but it can’t be the model of tomorrow.

And this whole shift that AXA is trying to embody by shifting from payer to partner, as we call it internally, is a paradigm shift to some extent, where we really want to strengthen the relationship with clients who rely on AXA, not only for protection, but also for prevention. Doesn’t change the fact that insurance will still be around and we will still need insurance cover, but it can’t be the sole answer to the risks.

Pete Miller [06:29]: Seems like it gives you a more comprehensive relationship with your customer, right? So, yeah, you mentioned a little bit about AI. So what can your AI predict that traditional insurance models are missing? And can you give us kind of an example of that?

Pierre du Rostu [06:43]: You know, the way insurance works and has worked forever is by looking at the past, right? This is how we use stochastic models and basically we look at average 20 years, average 50 years, average 100 years. And it works really well until the point where risks change.

And clearly, if you take climate change, for instance, what is the poster child of changing risks, you can’t anymore rely on historical data only.

And this is where AI, this is where new sensors such as geospatial data, for instance, comes into the equation. AI integrates live data, combining historical patterns of course, but as well using new data sets such as temperature, the presence of vegetation, the density, the winds and so on and so on, to create actual ground truth estimate of risks.

So, if we take a practical example, we’ve worked quite a lot on wildfires, and to address this challenge of wildfire, we use real time satellite data that we blend with 20 other factors such as topographic insights, temperature, density of vegetation, presence of human factor, et cetera. And we blend that so that in real time we have an exposure against risks that is, for instance, used by several firefighters companies across Europe, and they use that to prepare more effectively their operation and being faster at reacting to save property and lives.

Pete Miller [08:14]: So this is really forward looking. Where do you think this is all heading? What does the future look like when everyone has access to this kind of risk prediction?

Pierre du Rostu [08:22]: Well, it might be a bit overused, but I think we are ahead of a paradigm shift. I think it’s completely redefining the whole industry, maybe not tomorrow, but likely the day after tomorrow. We are moving from curative medicine to preventive medicine, so I go to the doctor when I’m sick, but instead of that, I’m going to the doctor in order not to become sick. And I think this is why our industry is changing. Risk management becomes proactive, simple, accessible to all lines of business.

Most of the organization will be protected, fewer disasters will happen, and insurance evolves into a service that not only saves the balance sheet of our clients, but also support sustainability and resilience of the companies, of the individuals and of the entire society. And we know it works.

There have been some studies done by the central bank that say that if you spend $1 on prevention, it saves up to $7 in reparation. So, we know there is a business case for our industry. And we know there is a business case for our overall society, and this is why we have to go that way, and this is why I think it’s so exciting to see this starting shift again and this revolution across our traditional industry.

Pete Miller [09:45]: I agree, Pierre, this is like amazing, you guys have certainly done an amazing job. So, is there anything else that you would like people to know?

Pierre du Rostu [09:55]: Yeah, so you know, I could probably talk forever about it, so this really excites me, and what’s really exciting as well is to test and learn with clients to discover together and to build solutions together.

And I think the answer can be only digital, and that’s why we work very tightly as well with our prevention engineering arms, and we can send boots on the ground to complement the digital angle. But digital gives us the ability to scale and to support clients globally, which is really, really new and really exciting for us because we know we can have an impact.

Last, but not least, we know where we are good at, but we can’t be good at everything. And this is why we chose the platform model, because around us we gathered an ecosystem of partners, some very niche startups, some very large companies who have part of the answer. And our goal is to orchestrate those answers into a comprehensive and meaningful value proposition to our users so that together we solve this challenge. It’s not a matter of AXA alone solving it, but AXA can help structuring the offer and the answer to those challenges.

Pete Miller [11:05]: Great, well, merci beaucoup, Pierre.

Pierre du Rostu [11:09]: With pleasure.