Is AI in Insurance Due Diligence worth the hype?

In 2023, investment in AI start-ups surged to nearly $50 billion, highlighting a growing enthusiasm for the technology's potential. OpenAI's CEO, Sam Altman, describes AI as a “super-competent colleague” that could intimately understand every facet of your life—an idea that both excites and unsettles. Yet, with tech company valuations in decline, some question whether the AI bubble is already deflating. As we reflect on previous tech cycles of boom and bust, we must ask: Is AI the “garlic bread” of innovation, set to enrich our work-life balance?

Hannah Fry, in her book Hello World, compellingly addresses the biases baked into AI systems. When AI is trained on datasets from past decisions, such as in sentencing, it tends to reproduce these biases rather than offer a perfect, impartial perspective. The phrase “garbage in, garbage out” still rings true. Fry even proposes a “magic” test to evaluate AI claims: substitute technical terms with the word “magic.” If the sentence makes as much sense—or seems just as plausible—then it's probably nonsense.

AI does hold enormous promise, particularly in scientific domains like physics and biology, where pattern recognition can be life-saving. It could revolutionize medical fields by transforming MRI analysis or cancer diagnostics. Businesses, too, are being sold on AI’s potential to drive efficiency and profitability, though often after an initial, hefty financial investment to see how AI might deliver on these promises—a proposition that invites skepticism.

Despite advancements in technology, many business owners still prefer the personalized service of brokers, especially in the world of online insurance, where technology investments have soared over the past 25 years. This is because complex business insurance needs rarely fit neatly into standardized, automated solutions.

At Vista, we’ve successfully closed over 500 deals in the past decade, ranging from tech start-ups to large multinational firms. The job demands thorough research and analysis. While AI could potentially streamline insurance due diligence—boosting efficiency, accuracy, and decision-making—it's crucial to integrate AI wisely. As brokers and entrepreneurs, we recognize the necessity of adapting to stay ahead, yet we’re mindful of AI’s limitations.

Key Opportunities and Challenges in AI Integration:

• Streamlining Data Collection and Analysis: AI can efficiently gather data, but the quality of its output depends on the quality of the data it's trained on. Inaccurate data can lead to flawed results and poor decisions.
• Enhancing Risk Assessment: AI offers deeper insights into risks, but overreliance on it may lead to overlooking human expertise. It should complement, not replace, human judgment.
• Automating Routine Tasks: Robotic Process Automation (RPA) can minimize errors and improve consistency, but system failures can have significant operational impacts.
• Improving Fraud Detection: AI can enhance fraud detection but still struggles with false positives and negatives.
• Enhancing Decision-Making: While AI delivers data-driven insights, biases in AI models and the necessity of human oversight remain pressing concerns.

Conclusion

AI has transformative potential for insurance due diligence, from data gathering to risk analysis and decision-making. Yet, we must approach the hype with caution. At Vista, we’re in the early stages of investing in AI to improve our services where possible, but we remain committed to the irreplaceable value of human expertise and intuition.