ZestyAI Expands Z-WATER™ Footprint with Regulatory Approval in Minnesota and Maryland

Property insurers in Minnesota and Maryland can now integrate AI-driven insights for non-weather water risk into their formal rating strategies. With regulators accepting ZestyAI’s Z-WATER model, carriers gain a more granular method to price a peril that accounts for $15 billion in annual losses, moving beyond traditional, less accurate territory-based underwriting metrics.

What was announced

The acceptance of Z-WATER in Minnesota and Maryland marks a significant expansion for the model, which is now approved for use in more than 20 states. The platform targets "non-weather water" risk—a category often overlooked compared to catastrophes but one that represents the fourth-costliest peril in homeowners insurance. These losses, totaling over $15 billion annually across more than one million claims, typically arise from internal failures like burst pipes, plumbing deterioration, and appliance malfunctions.

Traditional underwriting for these risks has historically relied on broad geographic territories and the age of the property, which offer limited insight into individual home susceptibility. ZestyAI’s model instead utilizes computer vision and aerial imagery alongside property characteristics, permitting history, localized climatology, and infrastructure context to assess risk at the individual property level. Trained and validated using insurer loss data, the company claims the model provides an 18x lift in risk segmentation compared to legacy territory- and age-based methods.

The move is part of a larger regulatory push by ZestyAI. As of September 2026, the company has secured more than 500 regulatory approvals across 317 carrier rate and rule filings. These approvals span 44 states and cover a suite of models including wildfire, severe storms, hail, wind, and non-weather fire, in addition to property intelligence.

The companies involved

ZestyAI operates as a risk decision platform specifically designed for the insurance industry, providing AI-powered climate and property risk analytics. Based in San Francisco, the company focuses on helping carriers, MGAs, and reinsurers refine their underwriting and rating processes through high-resolution data and machine learning.

The company’s portfolio addresses a wide range of perils that have become increasingly difficult for the industry to model using historical data alone. By leveraging computer vision and deep learning on aerial imagery, ZestyAI provides insights into property-specific vulnerabilities that are often invisible to traditional data providers. This includes everything from roof condition and vegetation overhanging a structure to the specific infrastructure risks associated with internal water damage. With nearly 60 reports on the company’s activities within the FF News network, ZestyAI has established itself as a prominent player in the North American insurtech ecosystem, particularly as carriers face mounting pressure from loss ratios and the need for more sophisticated pricing models in a volatile climate and economic environment.

What FF News has reported before

FF News has followed ZestyAI’s rapid expansion and its impact on carrier operations closely. Recently, the publication reported on how WEA Member Benefits Partners with ZestyAI to Transform Homeowners Underwriting with AI Insights. This followed news of a deepened relationship with a major regional player, as Donegal Insurance Group Expands ZestyAI Partnership for Enhanced Property Risk Intelligence.

In October, the publication highlighted a significant milestone when ZestyAI Hits 500 Regulatory Approvals for AI Property Risk Models Across 44 States. Additionally, FF News covered the company’s broader industry warnings, noting that ZestyAI Warns Insurers Are Paying an “Uncertainty Premium” Due to Critical Property Data Gaps, emphasizing the financial cost of relying on outdated property information.

What this means from an insurtech perspective

This expansion into Minnesota and Maryland signals a shift in how regulators view "black box" AI models. For years, the primary barrier to AI adoption in personal lines was the rigorous "actuarial justification" required by state departments of insurance. By securing these approvals, ZestyAI is effectively de-risking the adoption of AI for carriers. The 18x segmentation lift puts immense pressure on incumbent carriers still using territory-based pricing; those who fail to adopt property-level insights risk adverse selection, as more sophisticated competitors cherry-pick the best risks. For brokers and MGAs, this level of precision could lead to more stable capacity, as reinsurers gain confidence in the underlying data used to price high-frequency, non-catastrophe claims.