Z-WATER™ Approved in Five More States as Non-Weather Water Losses Hit Record Severity

Regulators in Utah, Colorado, Tennessee, Missouri, and Montana have accepted ZestyAI's AI-powered non-weather water risk model for use in carrier rate and rule filings, bringing the model's total approved footprint to 12 states nationwide.

Now the fourth-costliest peril in homeowners insurance, non-weather water drives more than $15 billion in annual losses across over 1 million claims, with average claim size exceeding $15,000. Routine failures like burst pipes and hidden leaks are now producing catastrophe-scale losses that surpass hurricanes in severity, yet the peril is difficult to model using traditional rating tools—which rely on territory-level or age-based proxies that overlook the property-specific factors driving interior water losses.

Using verified insurer loss data, Z-WATER™ applies computer vision to aerial imagery and incorporates property-level data, permitting history, localized climatology, and infrastructure context to capture the property-specific drivers of interior water losses. By modeling how these variables interact, Z-WATER predicts both the frequency and severity of non-weather water claims with 18× lift in risk segmentation compared to traditional territory- and age-based models.

"Non-weather water losses place real pressure on carriers' books, but they're also highly preventable when you understand where the risks actually lie," said Bryan Rehor, Director of Regulatory Strategy at ZestyAI.

"Z-WATER helps insurers pinpoint those vulnerabilities at the property level and price them appropriately, while meeting regulators' expectations for clarity and fairness."

These approvals add to ZestyAI's broader regulatory momentum. The ZestyAI platform — spanning wildfire, hail, wind, severe convective storm, non-weather water, and property and roof intelligence — has secured more than 200 regulatory approvals nationwide.

Carrier adoption and regulatory acceptance of AI rating models are accelerating in parallel, as the industry moves away from territory- and age-based proxies toward property-specific analytics.