2026-08-06 · Source: Insurance Business UK
Summary in 3 Points • Insurance firms may fail with agentic AI if treated as software rollout • Ramada points to the importance of orchestration and contextualisation over raw model intelligence • Insurers achieve significant savings and accuracy improvements with well-deployed agentic AI systems --- Dr. Magdalena Ramada Sarasola from WTW stresses that insurance firms may not realise the financial benefits of agentic AI if they approach it merely as a software rollout. She argues that the design and deployment of AI systems, including orchestration and contextualisation, are crucial for success. Insurers have seen significant improvements, such as $13 million in annual savings and 95% FNOL decision accuracy, when these systems are embedded in core workflows. However, the challenge remains in converting human expertise into system-compatible knowledge. Ramada also notes that while AI can enhance productivity, it should not be seen as a replacement for human expertise, especially in regulated decision-making processes.
For the London insurance market, the integration of agentic AI could lead to enhanced underwriting and claims processing efficiencies, potentially lowering operational costs. However, the market must approach AI as a transformative tool rather than a simple software upgrade. Insurers may need to invest in redesigning workflows and training staff to work effectively with AI systems. Additionally, the focus on maintaining human expertise alongside AI capabilities suggests that workforce reductions may not be advisable, as human judgment remains critical in complex decision-making scenarios.