The Federal Circuit and Family Court of Australia has issued an AI practice direction that applies to every proceeding in both divisions of the Court.
Effective from 29 May 2026, the direction permits the use of generative AI but requires Court users to verify outputs, protect confidential information and remain accountable for material filed or relied upon in proceedings. Should a lawyer choose to use AI, such use must remain consistent with their broader professional obligations.
“None of this should be a surprise to legal practitioners,” Wenee Yap explains, Director of AI Enablement (APAC) at StratAlliance Global, Professional Fellow at UTS Law, and Adjunct Lecturer in Technology & Innovation at the College of Law.
“The Chief Justice is simply issuing a practice direction that makes what should be common sense explicit, and it’s consistent with what we’ve seen from other Court directives on GenAI, and cases concerning lawyers disciplined for misuse of GenAI,” Wenee says. “This directive applies to all litigants, whether they’re self-represented or represented by lawyers, as well as expert witnesses and any other stakeholders involved in Court proceedings.”
“What’s great is that it recognises GenAI has become part of the reality of everyday legal practice, and it’s effectively seeking to regulate its use in Court proceedings,” Wenee explains. “This is exactly what we want. After all, if AI helps you prepare a submission, you still need to demonstrate a factual basis for every claim, citation or argument. You needed to do this before GenAI, and of course, you should need to do it if you use GenAI.”
Six principles to govern AI use
The Court takes a principles-based approach to regulated AI use, which Wenee says is consistent with best practice.
“Responsible AI is grounded in regulating the harm of AI use, rather than the technology,” she explains. “We’ve seen similar approaches in other fields; for example, when regulating the environmental impact acid rain, polluters were held accountable for the harms caused by their activity, some of which resulted in acid rain and the consequential depletion of the natural environment. It’s an approach that works. In that case, this harms-based approach to regulation helped restore ecosystems to better health.”
The Court set forth the following six principles to govern AI use in Court proceedings. According to the Practice Direction, these are:
- Integrity: Court users must only use AI in a way that is consistent with maintaining the integrity of Courts, and the important principles of fairness, accuracy, truth, proportionality and equality before the law.
- Accountability: Court users are accountable for the extent to which they choose to use AI, and the tools or systems they choose to use. Court users should be prepared to account to the Courts, if requested, about the way in which they have used AI, and provide evidence that it has been used responsibly, lawfully and in accordance with their obligations to the Courts.
- Accuracy: Court users must ensure that all information they rely on in Court proceedings is accurate.
- Confidentiality and data security: Court users must maintain confidentiality of information where that is required (for example by legislation, order of a Court or other duty) and understand the risks that use of AI can pose to the confidentiality of sensitive information.
- Safety: Court users must not use AI if there is a risk it will compromise the safety of any other Court user or child. This could include inputting information into a GenAI tool where confidentiality is not assured, and if disclosed, could impact a person’s safety (such as the disclosure of a protected residential address or medical records).
- Education and understanding: Like any tool or system, responsible use of AI requires awareness and education. Lawyers should undertake professional education on AI commensurate with the extent they intend to incorporate it into their practice. Court users must understand the strengths, weaknesses, risks and limitations of AI and especially GenAI before using it. The Courts will not accept a lack of understanding as justification for inadvertence in the use of AI.
How Should You Comply with the FCFCOA Practice Direction?
The challenge for law firms is turning policy into practice. Wenee works with major national, mid-tier and boutique firms on Responsible AI adoption and governance, and has delivered large-scale AI literacy training through the College of Law, UTS, and Legal Aid NSW.
“Having an AI policy sitting in a folder, or sent to staff via an email announcement, doesn’t amount to compliance. Lawyers need to understand what the rules set forth in documents like the FCFCOA Practice Direction mean for their everyday work. They need to know how to verify AI outputs, when and how to use AI, and which risks they need to manage.
“I often hear firms say, ‘It's fine, the software provider offers training.’ Vendor training has its place, but it's product training. It's designed to help you get more out of the tool, not to help you meet your obligations to the Court, your clients and your regulator. It won't tell you what needs to be verified, when a use case needs extra scrutiny, or how a practice direction like this applies to your work,” she said.
It’s no secret that AI adoption has outpaced governance Wenee observes.
“Regulators such as APRA and ASIC have already put organisations on notice, issuing letters to compel organisations to take appropriate steps to ensure they understand how AI is being used within their organisations, what risks are present, and how they are being mitigated. This is all occurring against a global context of AI agents driving hacks across trusted institutions like the Bank of America and Japan Rail.”
It’s a situation that requires lawyers and law firms to meet AI risk where it stands.
“Meaningful compliance needs strong governance, clear accountability, appropriate controls and ongoing education that is specific to the legal profession,” she says. “You can’t just say, ‘don’t use AI.’ Shadow AI use is so widespread that you can guarantee AI is already being used by your lawyers, and probably with varying levels of risk mitigation and irresponsibility. Good governance, properly operationalised, can enable AI to be used effectively, with appropriate safeguards, across an organisation, especially if you run AI pilots mapped to clear ROI.”