The Role of Actuaries in AI and AI Governance

Artificial intelligence is no longer a future risk for the insurance industry, it is a present operational reality. Yet the governance frameworks designed for classical models are structurally inadequate for AI systems that operate continuously at scale, behave adversarially, and propagate errors across fragmented value chains. This presentation makes three arguments. First, that the risk characteristics of modern AI map more closely onto cybersecurity governance paradigms than onto traditional model risk frameworks. Second, that AI governance responsibility is structurally fragmented, cutting across multiple internal functions and multiple regulatory regimes simultaneously, with no single framework and no single organisational actor currently owning the full picture. Third, that the actuarial profession, trained in model validation, embedded in the regulatory architecture through a signed and auditable opinion, and positioned across the full insurance value chain, is a natural candidate to bridge this gap. The presentation concludes with an honest assessment of where the profession stands today, acknowledging both the genuine strengths actuaries bring to AI governance and the gaps in curricula, risk literacy, and European professional coordination that must be closed. The central argument is not that actuaries are uniquely qualified. It is that they are uniquely positioned, and that the difference is the profession's responsibility to address.