Commonwealth Bank says it has deployed an “agentic AI” system intended to detect emerging fraud and scam patterns in transaction and payments data and propose new detection rules to help intercept them.
The bank said the work forms part of its $1 billion annual spend aimed at protecting customers from fraud, scams, cyber threats and financial crime.
CommBank said its existing fraud protection systems monitor more than 80 million signals each day, spanning transactions, card and online payments, and interactions with digital banking channels. The new AI agent is designed to assess suspicious patterns, analyse context and suggest updates to detection rules.
Executive general manager of fraud and scams James Roberts said the system is intended to speed up responses to evolving threats. “When suspicious patterns are identified, the system quickly assesses their severity, analyses context, and proposes new detection rules to help intercept them. The new agent goes beyond traditional AI by not only rapidly identifying new threats but also determining how it can seek to disrupt them,” he said.
Roberts said proposed rules are reviewed and approved by CommBank’s fraud analytics team before being implemented, in a human-in-the-loop process.
CommBank said it processes more than 20 million payments a day on average and sends more than 40,000 proactive warning alerts on average to customers via the CommBank app. The bank said its fraud detection technology helped reduce fraud losses by more than 20% in the first half of the 2026 financial year compared with the first half of the 2025 financial year.
The bank said the AI agent has contributed to developing or updating three quarters of its card fraud rules.
CommBank said the system was developed by its in-house data science and engineering teams in three months. It said the system runs on Snowflake’s data cloud and is supported by the bank’s cloud-based core banking platform.
The announcement comes as banks and payments providers increasingly look to automation and machine learning to respond to fast-changing scam typologies, while balancing speed with governance and accountability for decisions that can block or delay customer transactions.
