Insurance
Home›Insurance›Reinsurance›AI-informed catastrophe models aim to close storm pric…
AI-informed catastrophe models aim to close storm pricing gap at renewal
Severe convective storms drove $61 billion in global insured losses in 2025, and the US share was $51 billion, per Aon and Triple-I.
Severe convective storms have surpassed hurricanes as the costliest insured peril, pushing insurers to scrutinize whether the catastrophe models used for pricing are current enough for renewal, according to Insurance Business.
Aon data cited by the outlet shows SCS generated $61 billion in global insured losses in 2025, while the Insurance Information Institute, or Triple-I, estimates the US share at $51 billion, marking a third consecutive year above $50 billion.
Insurance Business reports that Karen Clark & Company, or KCC, argues traditional statistical cat models are poorly suited to SCS risk because the events are highly localized and can develop rapidly, making extrapolation from historical records unreliable.
The outlet says KCC’s white paper points to AI-informed model approaches as a next step, including models that ingest more than 30 gigabytes of daily satellite, radar, and weather data, and it adds that carriers using older statistical models may take a less current view, leading to uneven pricing in a softening property market.