Suncorp to have AI agents in insurance claims process as soon as this month - iTnews

Title

Suncorp to have AI agents in insurance claims process as soon as this month - iTnews

Content

Suncorp Group will deploy five AI agents handling sub-processes of insurance claims processing by the end of June 2026, marking the Australian insurer's first production agentic workflow. "The first set of ones, about five of them, are expected to go into production later this month," AI practice executive Kranthi Nekkalapu said at the Databricks Data+AI Summit in San Francisco. He framed the rollout as the first slice of a broader strategy: "Where we are right now is we are working through automating our claims processes using agents. Now, this is where everybody naturally wants to progress." Claims processing has long been an automation target because, despite its complexity, it follows a definable workflow: intake (FNOL or first notice of loss), triage and assessment, fulfilment, and settlement. "Claims processes are inherently very complex, but at a high level, they are workflows," Nekkalapu said. "[In an agentic world] the workflow remains the same, but within that, there could be several tasks. These tasks themselves could be done by agents." The deployment has progressed in two phases. Voice agents handling the initial intake of claims have already entered production: "We have got voice agents which went into production very recently. They do the initial intake of the claim in some cases, and if it becomes more complex, then pass on to humans." The five agents going live in late June 2026 will operate across the lifecycle, including two agents in the FNOL stage for capturing, validating, classifying, and routing the claim, plus agents for vendor dispatch and settlement calculation, with a "system of agents" performing the coverage check during triage. The technical architecture uses a business process model and notation (BPMN) workflow orchestrated with UiPath, with each task potentially handled by a single agent, a system of agents, or nested workflows. Observability is built on Databricks using OpenTelemetry (OTel) collectors ingesting traces in real time, producing role-specific dashboards for leadership, the safety team, business users, and engineers, with a Databricks Genie Code-powered chatbot for self-service debugging. Suncorp has assembled a controls library of approximately 100 controls across multiple AI application types and risk categories, with mandatory human review for all rejected claims. Suncorp is developing complex agents capable of determining claim coverage with 99 percent accuracy by analysing loss causes in scenarios such as storm damage with cascading effects ("a tree fell on the retaining wall, damaging a pipe that's running next to the wall, so that water seeped into a garage and damaged all the carpets"). The five-agent June 2026 deployment is a deliberate sub-process-by-sub-process approach ("We haven't automated the full end-to-end process, but there are sub-processes within that which we are automating at the moment") and serves as a foundation for broader agentic automation across the workflow.