Back to Articles
CBA to Take AI Orchestration Agent Beyond Its Retail Bank

iTnews

ENRICHED

Details

Date Published
9 July 2026
Priority Score
2
Australian
Yes
Created
10 July 2026, 06:00 am

Authors (1)

Description

Connecting customers to appropriate support.

Summary

Commonwealth Bank of Australia (CBA) is expanding its AI orchestration agent across its entire enterprise following successful implementation in retail customer service. Co-developed with Microsoft, the system utilizes 'intelligent routing' to triage interactions between conversational AI and human specialists based on context and sensitivity. While focusing on operational efficiency rather than catastrophic risks, the move highlights the maturation of multi-agentic workflows at major financial institutions. This development underscores Australia's rapid adoption of frontier AI orchestration capabilities within critical infrastructure and regulated sectors.

Body

Commonwealth Bank is set to take a central AI orchestration agent used in retail customer support and apply it to other areas of the bank. Details of the agent, which essentially triages inbound customer service requests to get them resolved faster, are revealed in a new case study of the technology published by Microsoft. The orchestration agent was first briefly showcased at the Microsoft AI Tour event in Sydney earlier this year, however the public case study is the first time that the project has actually been detailed. The AI orchestration agent for retail support is the result of two years of co-development by CBA and Microsoft. It is built on Copilot Studio, Dynamics 365 and Microsoft Foundry, the latter of which the vendor describes as an “AI app and agent factory”. The AI orchestration agent “interprets a customer’s intent and dynamically routes it to the most appropriate responder, whether that is another AI capability or a human specialist,” according to the case study, which adds that "the routing is intelligent and context-driven.” This means a straightforward question may be directed to be handled by conversational AI, drawing on publicly available information, whereas a sensitive conversation can be handed off to a human specialist, with AI remaining in the background “providing conversation summaries and potential responses”. General manager of assisted customer experiences Rachel Round said the system had produced “a step-change in how effectively customer enquiries are being resolved through [the bank’s] digital and messaging channels”. “In May 2026, approximately 84.6 percent of self-service messaging interactions were resolved end-to-end in the messaging channel,” she said. The case study states that the orchestration layer supports “future extensions into voice bots, multi-agentic workflows and, ultimately, enterprise-wide conversational banking.” It indicates that expanded use of the technology outside of the retail banking domain is now on the cards. “With the foundation now in place, CommBank is eyeing the next frontier to extend the platform across the entire bank,” the case study states.