With cyber being one of the top emerging risks for the re/insurance industry, some believe there are still many challenges to overcome, one of them being the creation of models to make it insurable, as it is argued there is not enough data yet to support them.
Data and analytics firm Russell Group believes this is not the case, as James Banasik, Head of Connected Risk commented during its press briefing at the 2023 Rendezvous in Monte Carlo.
“Some might argue that we don’t have enough data in terms of longevity. But we do have some data, and if you start to think about the data that we’ve got as a sign of what could happen, about things to come, then you start to see it,” Banasik notes.
Adding: “Like for example how a modern construction company could be impacted, in a different way to a bank, or a school or infrastructure, where it might be and just something that some of the similarities and some of the sort of cross-pollination of that data.
“Then you actually start to build a picture of what the future looks like, and you should effectively create these assumptions or scenarios that are based on this data.”
What the Russell Group is creating is a learning system, “with the emergence of artificial intelligence (AI) and predictive analytics, gathered data can be fed back into their model and whatever was a good prediction for a certain scenario, it can do that again with another one,” Banasik explained.
“I think it’s just about taking the data and really understanding it. We need to understand the behaviour of the data, not taking the data with face value and saying that’s the answer. And when you use AI then you’re actually mimicking the behaviour on the data, not the data itself,” he said.
“When you start grouping data together and look at what it is not likely to happen then you end up with fewer outcomes. This way you can start to see the wood through the trees, then start to see that, actually, it feels like it is going to be something over here. Then you start to drill down and I think that’s something that others that just jump straight in and try and boil the ocean, have problems with and the get overwhelmed by the quantity of data, and often the lack of quality data,” Banasik concluded.
