While the potential for transformation with AI technology remains vast for the insurance companies, the current implementation is far from robust. It is a $450bn opportunity on the horizon, however, only 10% of insurance organisations have partially or fully deployed AI agents, and a further 20% are merely dipping their toes into pilot projects according to a new report by Capgemini Research Institute.
The new 116-page report Rise of agentic AI: How trust is the key to human-AI collaboration reveals that a significant portion of the industry—42% of the insurers—has yet to define a formal strategy for integrating AI agents into their operations. As excitement brews within the sector for what is termed Agentic AI, many firms still grapple with the early stages of adoption, revealing a pressing need for clear implementation roadmaps.
Despite the rocky terrain, there is optimism about the role AI could play in daily operations. Sales and customer service emerge as frontrunners, with 52% and 57% of insurers, respectively, anticipating that AI agents will manage at least one process daily in these areas. Yet, despite these lofty expectations, trust in AI technology remains disturbingly low. Only 4% of organizations report full confidence in AI agents, and overall trust has dipped from an average of 54% in 2024 to 47% in 2025.
Chief concerns identified by insurers highlight critical issues surrounding privacy, bias, and transparency, each garnering concern from half of the surveyed firms. As these risks loom large, the path forward for AI integration remains fraught with caution.
Looking ahead, 26% of insurers believe that AI agents will begin augmenting human work within the next one to three years, while 35% foresee these agents operating independently under human oversight. The demand for skilled personnel also rises sharply; 67% of employers indicate that programming and software development skills will be crucial, while 60% emphasize the importance of decision-making as a key soft skill.
Yet, the ethical landscape appears murky; only a mere 8% of insurance organisations have successfully embedded ethical AI principles into their operations, significantly trailing the global average of 14%. As the financial sector increasingly embraces digital transformation, one can’t help but wonder: will insurers be quick enough to catch up, or will they find themselves left in a behind-the-times conundrum?