The Australian Securities and Investments Commission has made banks’ use of artificial intelligence in customer-facing activities a formal supervisory priority for 2026–27.
ASIC says it will examine existing and proposed AI uses and their effects on customers. Reuters reports that the review will include AI-supported decision-making and lending processes, reflecting concern that automation may advance faster than consumer protections and governance controls. The regulator has previously warned financial institutions about cybersecurity risks associated with frontier AI.
The review does not announce new AI-specific legal duties. Banks remain subject to existing obligations governing responsible lending, fair treatment, privacy, security and financial-services conduct. Its significance lies in ASIC moving from general consultation and warnings to sector-specific supervisory scrutiny.
That shift matters beyond banking. It suggests Australian regulators will increasingly assess AI through the legal obligations attached to the underlying activity, rather than wait for a comprehensive AI statute. A lending recommendation does not become less regulated because an algorithm or agent produced it.
For lawyers advising financial institutions and technology suppliers, governance should connect each automated output to its data sources, decision rules, customer impact and accountable human reviewer. Institutions should also preserve reproducible records showing how adverse or high-consequence decisions were reached, tested and, where necessary, overridden.
Sources: ASIC supervisory priorities; Reuters.
Leave a Reply