I’m thrilled to sit down with Simon Glairy, a leading expert in insurance and Insurtech, whose deep knowledge of risk management and AI-driven risk assessment has positioned him at the forefront of this transformative field. In our conversation, we explore how artificial intelligence is revolutionizing underwriting, reshaping risk assessment, and redefining pricing models in the insurance industry. We also delve into the evolving role of human underwriters, the benefits and challenges of adopting AI, and what the future holds for this dynamic sector.
How has AI changed the traditional role of underwriters in the insurance industry?
AI has fundamentally transformed underwriting from a manual, time-intensive process to a streamlined, data-driven one. Historically, underwriters relied on paper applications, actuarial tables, and personal judgment, which could be slow and inconsistent. Now, AI automates routine tasks like data collection and initial risk scoring, allowing underwriters to shift their focus to more strategic decisions, such as handling complex cases or refining pricing strategies. It’s not about replacing people; it’s about empowering them to add greater value.
Can you walk us through how AI is enhancing risk assessment in insurance?
Absolutely. AI is redefining risk assessment by processing massive amounts of data in ways humans simply can’t. Predictive analytics, for instance, combines historical data with real-time factors like behavior or environmental changes to forecast claims more accurately. Tools like computer vision assess property or auto risks by analyzing images or drone footage, while natural language processing speeds up document reviews by extracting key information from unstructured text. Even IoT devices provide live data streams, enabling dynamic risk profiles. Together, these technologies create a much sharper picture of risk.
What’s driving the move from static pricing to dynamic risk models, and how does it impact policyholders?
The shift to dynamic risk models is all about relevance and fairness. Static pricing often relied on outdated assumptions, but AI allows for continuous updates based on fresh data. For example, usage-based insurance adjusts premiums in real time based on driving behavior, while climate-sensitive models tweak property insurance rates using current weather risks. This means policyholders get pricing that reflects their actual risk, not broad averages, which can feel more personalized and equitable.
What are some of the standout benefits AI brings to underwriting for both insurers and customers?
AI delivers a win-win for insurers and customers alike. For insurers, it speeds up application processing from days to minutes, boosts accuracy by analyzing thousands of variables, and cuts costs through operational efficiency. For customers, this translates to faster decisions, quicker policy issuance, and a smoother experience overall. When underwriting is more precise and efficient, it also helps reduce loss ratios for insurers, which can ultimately lead to more competitive premiums for customers.
How do you envision the role of human underwriters evolving alongside AI in the coming years?
I see human underwriters becoming interpreters and collaborators rather than just processors of data. By 2026, with AI handling repetitive tasks, underwriters will focus on validating AI outputs, making judgment calls in nuanced cases, and working with data scientists to refine models. Human insight remains critical, especially in complex scenarios where empathy and contextual understanding matter. It’s a hybrid model—AI provides precision, but humans bring the depth.
What are some of the biggest challenges insurance companies face when integrating AI into underwriting?
There are several hurdles to navigate. Data privacy is a huge concern—insurers must protect sensitive information and comply with regulations while using vast datasets. Then there’s model explainability; customers and regulators want to know how decisions are made, which can be tricky with complex algorithms. Bias in historical data is another issue, as it can perpetuate unfair outcomes if not addressed. Lastly, there’s a talent gap—underwriters need training to adapt to AI tools. It’s a lot to manage, but leading insurers are building strong governance frameworks to tackle these responsibly.
What is your forecast for the future of AI in insurance underwriting?
I’m incredibly optimistic about where this is headed. By the end of the decade, I expect fully automated underwriting for smaller policies, instant underwriting embedded in e-commerce or fintech platforms, and hyper-personalized policies based on real-time behavior and context. We’ll also see more collaborative ecosystems where insurers, data providers, and Insurtechs work together seamlessly. It’s a complete redefinition of underwriting, moving from reactive to predictive, and I believe those who balance tech with human insight will lead the way in building trust and efficiency.
