Tropical cyclones research data is inherently biased towards the United States a cat modeller told the ILS Bermuda Convergence 2025.

Jamie Rodney, the chief executive of Reask, based in California, said it was easy to find information about hurricanes in America but much more difficult to find it for areas such as Fiji, Australia or even southeast Asia.

Reask is a climate risk analytics firm founded in 2018, which provides high-resolution weather risk analytics and forecasting to help people understand the severity and frequency of extreme weather patterns.

“One of the limitations of cat models is that they are not built globally,” Mr Rodney said. He added that this was partly because the US was the best at capturing hurricane data.

“They fly planes into hurricanes,” he said. “They have satellites. The guys in Fiji are just really thinking about intensity and location every six hours.”

Mr Rodney said Fiji lacked the resources to do more.

“The US has the largest exposure so drives the most amount of capital,” he said.

“If you need to service the globe or make the world insurable, focusing on the United States is important. But it also creates a bias in the fact that you can’t service the US and the rest of the world with the same resources,” Mr Rodney explained.

His firm has dealt with these biases by creating their own data sources to fill the gaps.

“That allows us to position ourselves quite nicely to say we are defensible against all of these things,” Mr Rodney said.

Kent David, the principal of hazard science and analytics consulting at Cotality, said: “It is not a surprise that there is vastly different descriptions of risk worldwide. The markets are able to capture, accept and accommodate that.”

Cotality, also based in California, provides data, analytics, and workflow solutions for the property services industry, helping businesses find opportunities and streamline operations. They offer services such as property data analysis, climate risk modelling, energy efficiency assessments and software for property surveying.

Mr David said artificial intelligence was allowing cat modellers to ingest data more quickly and with greater confidence.

“The general public often thinks of chat bots when they hear ‘AI’, but in cat modelling AI is a powerful tool for deriving information from large data sets,” he said.

His firm builds their models on physical representations of weather phenomena.

“That is important,” he said. “However, AI is a great way to reduce cost and increase speed to market.”

He said there was tremendous benefit from AI across the economy, but particularly in the cat modelling space.