Agentic AI represents a “step change opportunity” for life insurers if they modernise operating models with “intent and discipline”, according to a Deloitte report.

The technology can give cost, speed and decision quality advantages, and use of AI is already demonstrating material improvements across claims, underwriting, policy servicing and corporate functions, it says.

“The strategic question is no longer whether agentic AI can create value, but whether insurers have the foundations required to realise that value at scale. The next decade of life insurance in Asia-Pacific will be defined by the organisations that modernise now.”

Deloitte says agentic AI could unlock new sources of value if the right foundations are in place.

But intelligent systems will not compensate for unclear processes, outdated architectures or fragmented data, and the consulting giant recommends strengthening data and AI governance and equipping teams with the skills and judgment required for an AI-enabled future.

“When these elements are aligned, insurers are able to move from intent to execution and establish the foundations.”

For many life insurers, legacy core technology remains a key constraint, but Deloitte says as agentic AI “raises the strategic stakes”, new architectural and execution approaches are enabling insurers to modernise incrementally, reduce risk and move with greater speed and confidence than was previously possible.

“The first insurers to modernise their foundations will define the operating models that others must follow,” it says.

Deloitte says six foundations are needed to scale agentic AI:

• Redesign work around measurable outcomes, explicit performance indicators, and well-defined decision boundaries.
• Redesign processes so whatever can be is fully automated and human effort is focused where judgement and empathy matter most.
• Build cloud-native architectures that allow agents to retrieve data, trigger actions and co-ordinate workflows.
• Embed governance for “trustworthy AI”, spanning data quality, accountability, explainability and oversight.
• Create better-governed data – especially critical in life insurance where sensitive personal data, long-term obligations and regulatory scrutiny converge.
• Keep human judgment central – combine autonomous functions with human empathy and judgment.

See the report here.