Insurance companies are embracing artificial intelligence (AI) to improve their underwriting processes, and the results are promising. A recent report from Capgemini indicates that 62% of executives believe AI is enhancing the quality of underwriting and helping to reduce fraud. Additionally, 43% of underwriters now trust and regularly use automated recommendations from predictive analytics tools.
Robert Klepper, chief underwriting officer at ICAT, shared insights on how AI is changing the way underwriters assess properties. He explained that underwriters now rely heavily on location-specific data. This data helps them evaluate everything from residential to commercial properties.
One of the biggest challenges, however, is not having enough data but rather managing too much of it. Klepper noted that underwriters often struggle with data overload, especially when they are working from a desk. The key is to find ways to organize this data effectively to gain valuable insights for decision-making.
ICAT is using a mix of data sources, including dynamic sources like physical inspections and sensor data, as well as passive sources like satellite imagery. Klepper emphasized the importance of aerial imagery, which provides detailed information about properties that inspectors might miss during on-site visits. This technology allows underwriters to analyze and interpret various risk factors with a level of detail that was unimaginable just a decade ago.
The ability to view a property’s condition over time is another significant advantage of this technology. Underwriters can track changes and assess risks associated with nearby hazards, such as chemical storage or tree overhangs. This capability gives underwriters a clearer picture of potential risks and allows them to make more informed decisions.
AI is also helping underwriters process large amounts of information more effectively. Klepper mentioned that natural language processing is one area where AI excels. It helps structure data so underwriters can prioritize what information is most relevant to their assessments.
With AI handling the heavy lifting of data management, underwriters can focus on strategic decision-making. This also enhances their ability to mitigate risks before they become claims. For example, sensors can detect water leaks or heat issues, allowing for timely intervention.
While technology is playing a crucial role in underwriting, Klepper stressed that human brokers remain essential in the risk transfer process. Brokers act as advisors and advocates for their clients, providing insights that help underwriters make better decisions. When brokers and underwriters share the same information, they can make more accurate assessments and pricing.
Klepper also highlighted some industry-wide challenges. The risks faced by insurers are becoming increasingly complex and dynamic. Factors like commodity prices, inflation, and supply chain issues add layers of difficulty to underwriting. Recent natural disasters, such as the California wildfires, have further complicated the landscape, leading to significant claims payouts.
In 2024, U.S. insurers paid out around $140 billion in claims related to natural disasters, with the total economic cost estimated at $320 billion. These events have revealed risks that are broader and more extreme than previously understood. Traditional models may handle some risks well, but others, like wildfires and flooding, require more detailed data.
Klepper believes the future of underwriting lies in effectively harnessing big data. By improving the models used to assess risks, insurers can make more accurate decisions and better reflect actual risks in their pricing. The integration of AI and data science is set to reshape the insurance industry, offering both challenges and opportunities for underwriters moving forward.