AI and Accountability: ASIC Puts Artificial Intelligence in the Regulatory Spotlight
ASIC’s announcement of its 2027 enforcement priorities ( 18 November 2026) marks an important development in regulating artificial intelligence in Australia. For the financial services sector, ASIC has expressly identified consumer harm arising from poorly implemented AI and automated systems as an enforcement priority.
ASIC Chair Sarah Court captured the issue succinctly:
“The technology may be new. The responsibility is not.”
The message is clear. Businesses cannot delegate their legal responsibilities to artificial intelligence. Adopting AI does not diminish existing statutory obligations or the accountability of directors and senior management.
ASIC will specifically scrutinise systems that:
· Generate misleading information;
· Make inappropriate automated decisions;
· Operate without adequate supervision; and
· Prevent consumers from accessing effective support or human review.
This has significant implications for businesses and organisations developing and deploying AI. AI governance must extend beyond technical performance to encompass legal compliance, risk management, human oversight and consumer protection.
In my view, businesses should ask not only what AI can do, but whether they can legally justify, explain, and defend what their AI systems actually do. Your policy documents and internal approvals need to reflect this. I suggest recording your AI output on a case-by-case basis, to the extent you can statistically demonstrate accuracy in explaining or defending what your AI system actually does.
Innovation and accountability are not mutually exclusive. Indeed, accountability is fundamental to building public confidence in artificial intelligence.
Whilst decision-making may be automated, legal responsibility cannot be automated away.
Dr Will Higgs
Founder and Principal of Epoch Legal

