- What's at Stake: Getting better reads on performance, risks and trends for insurance executives
- Key insight: Executives need to be more hands-on with agentic AI
Agentic AI changes how insurance company executives work, what's required from insurers' workforces and how insurers' workflows should be organized, said Matthew Madsen, global commercial insurance lead at Accenture.
Insurance company leaders must get hands-on with agentic AI, and agentic AI should be "anchored" by the executives, Madsen said. "Today, many executives receive information through monthly reporting cycles, manual management packs and lagging indicators," he said. "AI-enabled insurers move toward real-time operating visibility, where executives will interact with AI to understand portfolio performance, emerging risks, claims trends, broker behavior and operational bottlenecks."
The insurance industry needs executives who "are not just comfortable with but thrive with perpetual transformation and change," Madsen added.
Similarly, insurance company staff need to be able to challenge AI output, according to Madsen. This requires fluency in AI, understanding how to supervise AI and interpret what AI produces. As a result, AI is changing the roles people play in jobs with insurers.
"New hybrid roles are emerging," Madsen said. "For example, the insurance AI decision scientist, who uses insurance domain expertise to design and implement AI-based decision-making frameworks and models, or AI-fluent experts in complex judgement under ambiguity, agentic collaboration and oversight, and ethical reasoning."
Since agentic AI does more than only automating tasks, it provides an opportunity to redefine workflows, but work must be standardized and consolidated before doing so, Madsen said.
Agentic AI increases the need for human experts who know the business, are fluent in AI and have decision-making authority, he adds. "These experts are excellent at joint problem-solving," Madsen said. "They can bring together people from across the organization, and systemically hold together the end-to-end thinking."
As a result, insurers' operating models are likely to be built around agentic AI, but with people working alongside the AI. "In short, the operating model moves from a process-centric operation to a decision-centric operation," Madsen said.
Functions such as business strategy, distribution strategy, strategic relationships, negotiations, pricing complex risks and interpreting the nuances of regulation will remain in human hands, but may be augmented with AI, according to Madsen.
