Twenty years ago, early in my insurance career, a seasoned CEO shared an insight that has shaped my perspective ever since: two pillars distinguish insurance from the broader financial sector – underwriting and claims handling. These are the entrance and exit points of a client's journey with an insurer, and they define the integrity, trust, and value of the entire business.

Shortly after joining, I attended a training course led by one of the world's top three reinsurers. The very first lesson we were taught was both surprising and profound: the best claims handlers are not necessarily medical doctors.

This was not a dismissal of medical expertise. Doctors spend years mastering their field, and their knowledge is undeniably vast. But their lens is clinical. They tend to view claims purely from a medical perspective, while effective claims handling requires a multidimensional approach. It's a delicate dance between logic, empathy, and foresight.

Fast forward, today we face a parallel opportunity with artificial intelligence (AI). Just as doctors bring clinical precision that must be balanced with broader business judgement, AI brings analytical power that reaches its full potential when combined with human oversight and contextual understanding.

When AI makes an erroneous decision, who bears responsibility?

The high-stakes reality of automation

Consider this real-life case: a client submitted reimbursement for eye drops prescribed for dry eyes. The medical report, as required, included a diagnosis – myopia. The AI system flagged myopia as an excluded condition and automatically declined the claim. What it failed to recognise was that the treatment addressed dryness, not myopia.

The potential consequence? Loss of a €2 million contract with a highly profitable client.

This incident illustrates AI’s critical limitation: it processes data but struggles with contextual interpretation. It detects patterns but needs to learn more about the nuanced judgement that preserves both contractual integrity and client trust.

The accountability question

As our industry accelerates AI adoption, we must confront uncomfortable questions. When AI makes an erroneous decision, who bears responsibility? The claims examiner who relied on the system? The technology team that deployed it? Senior leadership who approved its implementation?

More provocatively: as AI systems grow more sophisticated, how do we ensure they're trained ethically? The same technology that can streamline legitimate claims processing could theoretically be engineered to exploit policy ambiguities or systematically favour certain outcomes. What governance frameworks are we establishing? Who audits the algorithms? Where does accountability ultimately rest?

Human creativity bridges the gap

AI tools have become indispensable in our daily work, yet even the most advanced platforms reveal limitations. Recently, a sophisticated translation tool struggled with handwritten Greek text. The solution? Team members manually rewrote the letters on tablets so the machine could interpret them. That's human ingenuity bridging the gap where automation falls short.

The trust imperative

In health and life insurance, empathy isn't optional – it's foundational. If we lose the human connection between client and provider, we risk reducing a contract of trust into a mere transaction.

AI can and should amplify our capabilities – accelerating routine processing, identifying patterns human eyes might miss, and freeing skilled professionals to focus on complex cases requiring judgement. But it cannot replace the discernment, compassion, and contextual understanding that define excellence in our profession.