Allianz has used agentic AI for the first time to better manage Australian food spoilage claims and is now looking to roll out the tech to other countries and product lines.

It says “Project Nemo” cut spoilt food claim processing times by 80% as AI agents carried out repetitive tasks.

The insurer is considering using agentic AI in other low-complexity, high-frequency areas such as travel delays, simple motor claims or property damage assessments.  

“We scoped it intentionally ... It was the first agentic AI use case of its kind at Allianz,” chief transformation officer Maria Janssen said.

The Australian pilot provides a blueprint for broader application, she says, showing agentic AI can “transform how insurers respond to everyday losses, especially during a crisis. This initiative ... sets the foundation for integrating AI across Allianz’s operations, enhancing both efficiency and decision-making processes.

“The long-term vision is clear: a global ecosystem of specialised AI agents working alongside human experts to deliver insurance that is faster, fairer and more responsive – without sacrificing accuracy or trust.”

Nemo uses seven task-specific AI agents to manage coverage checks, detect fraud and more. A human makes the final payout decision.  

For food spoilage claims below $500, processing time has been slashed from several days to hours. 

Allianz says insurers must handle larger claim volumes with greater speed and precision as climate change drives catastrophes.

It says Nemo is a “splendid example of how agentic AI can deliver meaningful impact”.