Artificial Intelligence in insurance: are actuaries ready for AI? An insight from Aon

Aon: more than just an insurance broker

The conversation begins with an overview of the broader context of Aon’s operations. Although the company is commonly associated with insurance brokerage, its role is increasingly moving beyond traditional intermediation. As Ka Hei Choi emphasizes, close collaboration with insurers and a deep understanding of market challenges have naturally led Aon to develop its own technological solutions.

One key example is PathWise—an actuarial modeling platform based on GPU architecture. This solution not only significantly speeds up calculations but also paves the way for a whole new scale of data analysis in life insurance.

Faster calculations aren’t enough

Tomasz Dąbkowski looks back to the early days of his career, when advances in hardware and software gradually accelerated actuarial processes. Faster models meant more time, but it quickly became clear that every improvement in performance raised new questions from managers and regulators. “If we can calculate more, let’s calculate more”—this is a mechanism the industry has known for years.

According to Ka Hei Choi, the problem is that traditional actuarial models were not designed with future scalability in mind. CPU limitations, high code rewriting costs, and the tight coupling of actuarial logic with technology mean that many organizations are stuck with architectures that are difficult to adapt to new realities.

Is AI technology mature yet?

The question of artificial intelligence’s maturity comes up directly in conversation, but the answer is not clear-cut. Ka Hei Choi suggests shifting the perspective: instead of asking whether AI is ready, it is worth asking whether actuaries and organizations are ready to use it.

Just as Excel once revolutionized the work of finance professionals, AI today has the potential to become the new standard. At the same time, the history of Excel shows that without proper rules, expertise, and oversight, technology can lead to chaos and “black boxes” that no one fully understands.

AI is like a sports car—you have to learn how to drive it first

A suggestive metaphor involving a Ferrari comes up in the conversation. Artificial intelligence offers immense power, but without the skills and a plan, it’s easy to spin out at the first turn. As Ka Hei Choi notes, AI isn’t about a single agent that “does everything.” The key is designing the entire architecture, in which various AI agents support humans at specific stages of the process.

“Instead of asking AI to build an actuarial model for us, we should consider what questions to ask and how to develop an appropriate step-by-step action plan,” the podcast guest emphasizes.

Productivity, but with responsibility

One of the most frequently raised arguments in favor of using AI is the dramatic increase in productivity. Tasks that previously took weeks or months can now be completed in a matter of hours. This is changing the way teams work and also business expectations.

At the same time, both speakers agree that responsibility for results cannot be shifted to algorithms. Models must be understandable, verifiable, and explainable, especially in the context of relations with regulators. Artificial intelligence is meant to support human decisions, not replace them.

Regulations and the human element

Regulatory pressure in insurance is inevitable, but—as Ka Hei Choi notes—dialogue with regulators remains a human interaction. That is why it is so important for AI-based solutions to be transparent and to allow for a clear explanation of how specific results were generated.

This approach requires thoughtful governance and clearly defined roles for humans and technology in the decision-making process.

The wave has already arrived

Toward the end of the conversation, the metaphor of a wave comes up. The technological wave has already arrived, and other sectors of the economy learned how to “surf” it long ago. In the actuarial field, changes are happening more slowly, but—as the interviewees emphasize—the question isn’t whether the wave will come, but whether the industry will be ready to embrace it.

By developing solutions such as PathWise and investing in AI, Aon aims to act as a catalyst for this change. “We need to put on our swimsuits and get in the water,” concludes Ka Hei Choi.