AI agents are quickly moving from novelty to practical business tools. They can book appointments, search systems, complete tasks, interact with platforms and make decisions with less direct human input. That makes them useful, but it also creates a new governance problem. A recent Australian report by The Guardian highlighted the issue clearly, if an AI agent causes harm, makes the wrong decision, misuses access or acts in a way nobody expected, the agent itself is not the o

APRA Member Therese McCarthy Hockey's remarks to the 2026 AFIA Risk Summit are a timely reminder that AI governance is no longer a “future issue” for financial institutions. AI is already being used in decision-making, customer-facing processes, fraud detection, claims handling, software development and cyber defence. But as adoption accelerates, so do the risks. APRA’s message is clear, organisations need to understand how AI is being used, where it sits in critical operatio

AI adoption is moving quickly, and Australian businesses are being encouraged to use AI safely, responsibly and with proper governance. The Australian Government, including the Department of Industry, Science and Resources, the National AI Centre and AI.gov.au, has released guidance and voluntary safety standards covering areas such as accountability, risk management, transparency, data governance, testing and human oversight. Mandatory guardrails for high-risk AI settings ar

AI is already being used across Australian businesses — often before proper rules, policies or risk controls are in place. The Australian Government’s AI adoption guidance is a timely reminder that businesses do not need to wait for a dedicated AI law before taking action. If AI is being used inside the business, there should be a clear framework around how it is used, who is responsible for it, and what risks need to be managed. The guidance focuses on six core areas: accoun