Secondary Perils Aren’t Secondary Anymore
As natural catastrophe losses continue to soar, capturing global attention, insurers are seeking new and better ways to assess and manage risk. At a recent NAMIC conference, Attila Toth, Founder and CEO of ZestyAI discussed the role of artificial intelligence (AI) and machine learning in mitigating property and climate risk. These cutting-edge technologies empower insurers to make smarter decisions by extracting valuable insights from unstructured data sources like imagery, building permits, climatology data, and historic loss data.
The insurance industry has a well-developed understanding of primary perils, which are large infrequent events that are well-monitored and modeled. However, for secondary perils, there isn't the same level of understanding, monitoring, or modeling. These more frequent but smaller events, like severe convective storms, are smaller in scale but have been increasing in frequency and severity, causing billions in damages. This year alone, the industry is expected to hit $50 billion in storm losses - an all-time record - and those losses are only increasing as inflation, population density, and climate change continue to drive costs up.
Gaining a Better Understanding of Risk with AI
ZestyAI uses the latest advancements in AI to extract insights that can be used to make smarter decisions about climate-driven losses. In order to harvest these insights, we deploy machine learning and deep learning models that, once trained, can extract insights from unstructured data sources such as imagery, building permits, climatology data, historic loss data, and more.
Machine learning and deep learning models are trained through a process similar to how humans learn, which is to present examples to the models and then have the models figure out or ‘learn’ how to make future predictions based on the information they have been trained on. At ZestyAI our models can tell the difference between a wood shake roof and a clay tile roof, they can identify solar panels and pools, and they can even determine the slope of a parcel of land or pick up a single missing roof shingle.
How Can AI Improve Risk Assessment and Mitigation?
AI has made it possible to understand individualized property risk and price that risk accordingly on a policy-by-policy basis. That means insurers can price fairly and according to risk, instead of creating blanket rates for all homes within a particular postal code or geography, regardless of each property’s individual risk profile. For example, the new severe convective storm modeling approach that ZestyAI and the IBHS collaborated on showed more than a 50X improvement in risk-splitting power vs. status quo methods.
Because AI has the ability to process and analyze vast amounts of data quickly and accurately, insurers are able to access precise information about a property instantly. Our models are also able to understand the complex relationship between variables and identify patterns and trends, which leads to improved risk modeling that adjusts and adapts as things like climate risk evolve. Improving the speed and accuracy of risk information is a benefit to both insurers and policyholders because the risk to any given property is able to be assessed and priced fairly, reducing losses while maintaining access to high-quality insurance coverage.
Additionally, ZestyAI models were built with comprehension and transparency in mind, so our climate risk models provide a risk score that is accompanied by the top modifiers that led to that score. This gives carriers the ability to communicate with policyholders on what specific actions can be taken to lower their property’s risk, such as repairing missing shingles or cutting down surrounding dry brush or overhanging vegetation.
Use Cases of ZestyAI Across the Insurance Value Chain
During the talk, Attila highlighted a variety of use cases for ZestyAI across the insurance value chain:
Superior New Business Underwriting
ZestyAI simplifies and improves the quoting process, allowing insurers to provide faster and more accurate quotes for their customers, reducing the need for time-consuming manual reviews and inspections.
Access to exclusive insights on property characteristics and risks for every property in the US enables insurers to tailor pricing and rate plans based on individual property risk profiles. This ensures fair and appropriate coverage for policyholders.
ZestyAI's advanced risk models enable insurers to optimize their existing portfolios by identifying the highest-risk properties. This allows actions to be taken prior to renewal to reduce loss ratio and churn.
Attila's talk emphasized the game-changing potential of AI in revolutionizing the insurance industry's approach to risk assessment and mitigation, particularly for severe convective storms. With climate risk continuing to change the insurance landscape, there is an enormous opportunity for insurers to turn the tide on losses by harnessing the power of these cutting-edge technologies. ZestyAI provides insurers with the tools they need to make informed decisions, reduce losses, and ensure fair and accurate pricing to foster a more sustainable insurance ecosystem for the future.