Insurer use of agentic AI creating ‘new generation of conduct risk’

Insurers are increasingly deploying autonomous artificial intelligence (AI) agents across the policy lifecycle and their growing reliance on the technology is introducing a “new generation of conduct risk”.

This is according to the latest briefing in the AI Impact Report series from professional services and technology firm Davies Group, released today (30 July 2026) and titled How AI agents are amplifying conduct risk in insurance.

The report highlighted that while existing regulatory frameworks already apply across business functions that may choose to utilise agentic AI, the evolving and sometimes unpredictable decision-making skills of said agents puts control and oversight into question.

Indeed, Davies argued that the introduction of agentic AI technology brought with it “implications for fairness, transparency, accountability and customer outcomes”, especially when systems operated with “limited visibility or insufficient controls”.

Accountability rules

Three key issues were identified across the current AI landscape – increased opacity and reduced explainability of decisions, reinforcement bias growing over time and agents deviating from guidelines when allowed to operate in an autonomous decision loop.

Davies added that firms must ensure AI-made decisions remain “explainable and reconstructable”, with comprehensive audit trails and input capturing, to avoid falling foul of accountability rules.

The report concluded: “The widespread deployment of AI agents marks a pivotal shift for insurers in the UK and Ireland – one that brings significant opportunity, but equally heightened regulatory, ethical and operational responsibility.

“As agentic systems become more autonomous, adaptive and embedded within core insurance processes, the challenge is no longer simply adopting the tech but implementing AI governance and compliance in a way that is demonstrably fair and resilient.”