As insurers increasingly deploy autonomous AI agents across customer interactions, underwriting, claims and operational processes, a new generation of conduct risks is emerging.
Unlike traditional AI models, agentic AI systems can make decisions, adapt their behaviour and act with a greater degree of independence, creating new challenges for governance, oversight and customer protection. This paper explores how AI agents are reshaping conduct risk across the insurance sector in the UK and Ireland. It examines the implications for fairness, transparency, accountability and customer outcomes, highlighting how risks can be amplified when AI systems operate with limited visibility or insufficient controls.
Drawing on current regulatory expectations, including Consumer Duty, GDPR and the Consumer Protection Code 2025, the paper explains why existing conduct and compliance obligations already apply to AI-driven decision-making. It also considers why these obligations
become more complex as AI systems evolve from static tools to autonomous agents capable of learning and adapting over time.
A key theme is the need for insurers to move beyond traditional, point-in-time compliance approaches. Effective AI governance requires continuous oversight across the entire AI lifecycle, from design and deployment through to monitoring, testing and ongoing assurance. Firms that fail to adopt this approach risk not only regulatory scrutiny but also erosion of customer trust.
Part of the Davies AI Impact Series, this report provides practical insight into one of the most significant risk and compliance challenges facing insurers today. It is designed to help leaders understand how AI agents are changing the conduct risk landscape and what steps they can take to deploy AI responsibly while meeting evolving regulatory expectations and protecting customer outcomes.
Be the first to comment