Insurers Embrace Agentic AI as Trust and Risk Take Center Stage

A new report from the Capgemini Research Institute finds insurers beginning to integrate agentic artificial intelligence (AI) into operations, despite lingering concerns over trust and accountability. The study forecasts up to $450 billion in economic value by 2028 from agentic AI and reveals that 20% of insurance firms are already testing use cases, with 12% implementing the technology at scale. However, just 4% of insurers report full trust in AI agents, underscoring the persistent gap between technical potential and practical confidence.

Agentic AI shows promise in enhancing underwriting, streamlining claims processing, and supporting customer service. Yet, the insurance sector—bound by regulation and high ethical standards—is moving cautiously. Trust in fully autonomous AI has dropped significantly, with many insurers favoring collaborative models where AI supports, but doesn’t replace, human roles. These hybrid systems are expected to boost both productivity and employee satisfaction.

Insurers are also eyeing new exposures and liabilities that AI agents could introduce. Emerging risks tied to decision-making algorithms, cyber vulnerabilities, and third-party liabilities are driving insurers to explore new types of coverage for algorithmic operations. Despite hesitancy, the report shows confidence tends to grow as organizations progress from pilot projects to implementation.

Wider adoption hinges on overcoming readiness challenges. Only a fifth of organizations claim to have mature AI infrastructure, and even fewer report strong data preparedness. Privacy, accountability, and explainability remain top-of-mind. Still, as more insurers move to scale their AI strategies, the sector is poised to redefine risk management and operational efficiency in the age of autonomous systems.