Setting the Stage
I spent more than three decades in the insurance industry at one of the largest P&C carriers, leading multiple waves of enterprise technology transformation. Artificial intelligence represents a fundamentally different shift. It is not just another modernization cycle. It is quickly becoming a competitive requirement for both organizations and the people who work within them.
Historically, P&C insurers have been viewed by the broader U.S. technology community as cautious adopters. That reputation is not entirely misplaced, but it reflects structural realities more than a lack of ambition. The industry has often favored proven paths over early experimentation, prioritizing stability, regulatory alignment, and long-term risk management.
That traditional “wait and see” posture now carries real consequences. With AI, and especially the rapid adoption of large language models, delay does not create safety. It creates distance from learning, talent development, and operational advantage. For many carriers, hesitation may prove more disruptive than early, disciplined, frequent scaled delivery.
The Industry Mindset May Be the Biggest Constraint
A recent Bain & Company report noted that most P&C carriers are still “dabbling” in generative AI. That may not surprise anyone inside the industry. The insurance industry rarely struggles with understanding technology. It often struggles to decide when risk is manageable enough to move at pace.
For decades, the industry has been trained to manage downside risk such as data exposure, regulatory pressure, brand protection, customer impact, and operational stability. Those instincts are not wrong and are part of what makes insurance durable. But in the AI era, they become the very force that impedes competitive learning and early value delivery.
Too many organizations are waiting for the perfect “home run” use case before committing. The irony is that home runs rarely appear before momentum exists. They emerge after many small real-world initiatives, customer-driven iterations, and operational wins. When leaders insist on a transformational breakthrough before starting, progress stalls, teams lose energy, and competitors learn faster. Meanwhile, modern AI adoption moves forward in other industries.
The biggest advantage for P&C carriers is not one massive deployment but rather a steady stream of high-velocity base hits. Small improvements across claims handling, document processing, underwriting analysis, and financial operations compound into real organizational capability. These opportunities are numerous, building muscle memory and attracting talent. Over time, they change how work gets done.
The real risk today is not experimenting too early. It is waiting too long, allowing learning gaps to become competitive disadvantages with real financial consequences.
Mindset Shift to Deliver Dozens of Small, High-Value Opportunities
Having lived through this many times while driving innovative change, I believe a mindset shift is needed. The path forward is not a dismissal of downside risk in AI, but rather a rebalancing of how those risks are weighed. The pace of progress should be reset around these considerations:
· Common risk frameworks cite nine risk categories (https://www.diligent.com/resources/blog/strategic-risk-examples). Over my career, these frameworks have provided strong macro guidance. One consideration is to rebalance attention toward mid-term competitive, financial, and operational risks. To achieve aspirational growth and customer goals, investment in mid-term value is essential. Too often, these investments are viewed as threatening today’s business rather than creating opportunities for tomorrow.
· This point of view may be perceived as suggesting that no “home run” AI initiatives should be pursued. My objective is not to discourage those investments but to emphasize the need for smaller production wins on the path to larger outcomes. Any larger idea can be broken down into smaller wins. For example, if a carrier wants to reimagine the claims process using AI extensively, that is a strong pursuit. A process like First Notice of Loss contains many components and steps. While working toward that reimagining, implementing comprehensive document ingestion for a critical step delivers near-term value while creating significant long-term benefits.
· Organizational culture shapes how risk management and compliance are administered. Prior to the LLM surge in late 2023, AI risk was often loosely governed within decentralized frameworks, while carriers focused heavily on model management, especially regulated pricing models. As LLM adoption accelerated, many organizations centralized AI risk to increase awareness and control. In my experience, most product development risk operated in a hybrid model, enabling solutions to be built close to the business. Organizations should recalibrate AI governance toward a hybrid approach to maintain competitive pace, combining centralized oversight for significant initiatives with faster pathways for lower-risk efforts.
Strategic Advantage Driven by the Pace of AI Learning
There is no cookie-cutter formula for driving business transformation with AI. The perspective shared here was shaped through incremental delivery and early learning. The industry has the talent and capability to lead what is, in my view, a fundamental shift. There are many paths to success. Please share your perspectives and counterpoints. Open debate is essential.