ZestyAI, a risk and decision intelligence platform for insurers, has launched Z-SPARK, an AI-powered model designed to predict non-weather fire risk at the individual property level.
The model analyses the conditions that influence fire ignition and spread, helping insurers identify structures most likely to generate significant losses.
Non-weather fire incidents remain a major source of insurance losses. In 2023, approximately $25bn in property losses were attributed to fires caused by sources such as electrical faults, appliances, heaters and outdoor grills.
Many insurers still assess fire exposure using neighbourhood-level averages and historical loss data. Z-SPARK instead evaluates risk at the property level by analysing factors including building materials, maintenance conditions, surrounding structures, local fire response capacity and climate influences.
The system uses machine learning models trained on millions of fire incidents and verified insurance claims to estimate both the probability of ignition and potential loss severity.
Insurers using the model can align pricing more closely with actual property risk, support straight-through processing for low-risk properties and focus underwriting attention on higher-risk exposures. The platform can also help insurers manage portfolio concentration risk and expand into challenging markets with improved risk visibility.
ZestyAI said Z-SPARK delivers up to 30 times greater risk differentiation compared with traditional territory-based fire models.
Kumar Dhuvur, Founder and Chief Product Officer, ZestyAI, said. “Non-weather fire is one of the most costly and least understood risks in property insurance. Two homes on the same street can have dramatically different fire risk depending on how they were built, maintained, and what surrounds them. Yet most insurers still evaluate that risk using community-level scores. Z-SPARK reveals those hidden differences so insurers can make decisions based on the actual risk at each property.”
The launch builds on ZestyAI’s existing catastrophe modelling capabilities, including its Z-FIRE wildfire risk model. The new tool extends the company’s property intelligence suite to everyday building fires, which represent one of the most frequent sources of insured property loss.
Z-SPARK forms part of ZestyAI’s broader platform for analysing property exposure, alongside models covering hail, wind, severe convective storms and water damage.
Keep up with all the latest FinTech news here
Copyright © 2026 FinTech Global
