BMA conference: P&C leaders say data, AI, tech are reshaping risk decisions
Data, technology and AI are moving from experimentation to execution across the property/casualty market, as carriers look for sharper insight, faster decisions and better ways to manage a more complex and interconnected risk landscape.
That was a central message from the session “P&C CEO Round Table: Managing Risks and Capital Solutions in an Evolving Global Landscape” at the 2026 BMA Forum, entitled “Charting the Course: Managing Risk and Complexity in a Rapidly Evolving Landscape”, which took place in Bermuda this week.
Moderated by Shanna Lespere (pictured), deputy CEO, Bermuda Monetary Authority, the panel featured Andy Marcell, CEO, Global Solutions, Aon; Jim Williamson, president and CEO, Everest; Lucy Pilko, CEO, Americas, Axa XL; and Nicolas Papadopoulo, CEO, Arch Capital Group.
Lespere set the tone by pointing to a market shaped by increasingly complex risk realities, and by a concern that technology is moving quickly enough to challenge both companies and regulators. From the outset, the discussion suggested that AI is no longer a future topic. It is already affecting underwriting, claims, operations and client engagement, even if the industry is still working out how to use it most effectively.
Pilko framed the issue first through the client lens, arguing that the total cost of risk is rising and that insurance is only one part of how companies respond. “The world is getting riskier,” she said, adding that carriers sit on a “mountain of data”, especially claims data, that should be used more actively to help clients manage resilience rather than simply buy cover.
At Axa XL, she said that has meant investing in risk engineering, digital engagement and tools around areas such as wildfire, as well as thinking more broadly about the risks clients themselves face from AI.
Pilko pointed to work on endorsements related to training large language models and said insurers have to understand the strategic priorities of their clients if they want to provide relevant solutions. In her view, technology should support better judgement rather than replace it. “Technology has to make us, help us, make decisions,” she said.
Williamson suggested the industry is now moving into a more practical phase of AI adoption. After a period dominated by proofs of concept and experimentation, he said the past 12 months have started to produce measurable business benefits. At Everest, the emphasis has been less on cutting headcount than on making underwriters more effective by allowing them to spend more time on the best opportunities.
He highlighted the company’s wholesale business in the US, where more than 90% of property and core general liability submissions are now fully automated from intake through to presentation for the underwriter. For Williamson, the value lies in helping experienced professionals focus their time where it matters most.
Claims, he added, may be an even more important area of change. AI is helping insurers analyse legal environments in more detail, identify problematic claims earlier and deploy resources more intelligently. In the context of social inflation, he said that can produce better outcomes for both clients and carriers. “The technology is playing a very powerful role right now in escalating claims, in managing fraud,” he said.
Papadopoulo picked up that theme from an underwriting perspective, but with a note of caution. He said the danger with some AI tools is that they can create a false sense of certainty. Models can narrow outcomes and offer high-conviction answers, but that does not necessarily mean they are producing the right strategic result for the company. Underwriters, in his view, still need to step back and ask whether a model-driven answer makes sense in the context of a broader portfolio.
That need for judgement, rather than pure automation, was echoed by Marcell, who brought in the broker perspective. He argued that the industry sometimes forgets what it actually sells: not certainty, but a structured transfer of uncertainty. If outcomes were fully predictable, he suggested, much of the sector would not exist in its current form.
For Marcell, AI can improve the quality of decisions and accelerate access to data, but it cannot eliminate the need for human judgement around volatility, structure and the range of possible outcomes. He described the role of the industry as measuring uncertainty and transferring it to counterparties, whether through traditional balance sheets, sidecars or other forms of capital. That is becoming more important, not less, as geopolitical shocks, climate change, cyber risk and social inflation widen the range of possible outcomes.
The panel also tied technology directly to operational performance. Pilko said the ability to present the right information to the right person at the right time is increasingly what determines responsiveness. At Axa XL, she described the goal as making teams “bionic” by giving them better intelligence at the point of decision-making.
Williamson linked that to culture and authority. At Everest, he said one guiding principle is that the person closest to the risk should be the one making the decision. Technology can enhance that model, but it does not replace it. Instead, it helps leaner teams operate with greater confidence and speed.
The conversation then moved to talent, where both opportunity and tension were evident. Lespere raised the concern that technology could outpace the ability of new entrants to build experience. Pilko responded by arguing that the industry needs to rethink how it develops people. She called for more “shoulder-to-shoulder apprenticeship”, with experienced managers articulating how and why they make decisions so that younger professionals can learn faster in a more AI-enabled environment.
Williamson took a slightly different view, suggesting the industry has already been through similar transitions before. Underwriters no longer learn many of the basic manual tasks that were once seen as essential, yet the profession has adapted. His conclusion was that younger professionals will learn what is necessary for the environment in which they operate.
Taken together, the panel painted a picture of a market in transition. Data, technology and AI are clearly becoming more embedded in how insurers and brokers operate, but not as substitutes for experience, relationships or underwriting judgement. Instead, the message from the roundtable was that the winners in 2026 and beyond will be those that use technology to sharpen human decisions, improve responsiveness and engage clients more intelligently across a risk landscape that is only becoming more complex.
Did you get value from this story? Sign up to our free newsletters and get stories like this sent straight to your inbox.
