Federal Court releases Use of Generative AI Practice Note: key guidance for using AI in proceedings
The Federal Court of Australia released its ‘Use of Generative Artificial Intelligence Practice Note’, – just yesterday – marking the court’s first comprehensive statement on the use of generative artificial intelligence (Gen AI) in proceedings before the court.
The Practice Note sets out the Federal Court’s expectations and guidance regarding the use of AI in connection with proceedings before the court and comes amid a steady stream of cases in Australia and internationally involving AI-generated ‘hallucinations’. It squarely addresses that the presentation of false or inaccurate information to the court is ‘unacceptable’.
Significantly, this Practice Note also warns that caution should be taken when entering confidential or privileged information into AI tools. In referencing the recent Victorian Law Reform Commission, ‘Artificial Intelligence in Victoria’s Courts and Tribunals’ Report (tabled in February 2026), particularly in relation to the risks of inadvertent waiver, the Practice Note represents a step forward in providing aligned guidance on the appropriate use of AI.
The Practice Note
The Practice Note recognises the potential benefits of Gen AI, including increased efficiency, reduced costs and improved access to justice, while placing equal emphasis on the risks posed by misuse of these tools in litigation.
This is offset by the requirement that whoever has prepared a court document must have first confirmed its reliability and veracity, to avoid presenting hallucinations or false evidence to the court.
It makes clear that any use of Gen AI must remain consistent with existing legal and professional obligations and not compromise the administration of justice or public confidence in the legal system. In so doing, the Practice Note requires that in some circumstances, the use of Gen AI must be disclosed to the court.
The Practice Note applies to all persons who appear before, or file documents with, the Federal Court.
Dealing with confidential, suppressed or private information
The Practice Note directly addresses the risks that arise when sensitive information is input into Gen AI, with an entire section included on this point.
Information provided to a Gen AI tool may become available to others, as users may not know where that information is stored, how it is used, or who will have access to it.
It is specifically cautioned that entry or use of privileged, confidential or private information must not occur in a way that is inconsistent with the obligations applying to the use of that information.
Users of Gen AI should carefully consider these restrictions before inputting this information, with this guidance directed particularly at ‘open’ or ‘public’ forms of Gen AI tools.
Alignment with the Victorian Law Reform Commission’s Report
The Practice Note also addresses the Victorian Law Reform Commission’s, ‘Artificial Intelligence in Victoria’s Courts and Tribunals: Report’, which we have previously discussed in our Smarter Lawcast podcast series, ‘AI and legal professional privilege’. That report marked the first inquiry by an Australian law reform body into the use of AI in courts and tribunals and made 30 recommendations to the judiciary and practitioners regarding the safe use of AI.
Following guidance from the report, it notes that the risk of inadvertent disclosure may be lower for tools that operate in ‘closed’ or ‘controlled’ environments and draws comparisons between ‘public’ and ‘closed’ AI. In so doing, a clear distinction is drawn between ‘open’ or ‘public’ Gen AI tools and ‘closed’, private or enterprise systems.
The heightened dangers associated with publicly accessible tools are also highlighted, particularly where data handling, retention and training practices may not be transparent or controllable by users. These risks align with those identified in the report, which emphasises caution on the confidentiality, privacy and privilege risks associated with the use of Gen AI – particularly where information is uploaded to public models and may be retained or reused outside the user’s knowledge or control.
The report further notes that unmanaged use of AI tools could undermine legal professional privilege and the protection of sensitive and confidential information and encourages adoption of its recommendations to ensure safe and responsible use of AI in courts and tribunals.
Only two months after the report was tabled in Parliament, the Federal Court’s Practice Note can be seen as an early adopter of part of this report. While there is no explicit formal adoption of the Commission’s formal recommendations, the Practice Note’s guidance around dealing with confidential and privileged information reflects, and seeks to address, the same underlying concerns.
AI use in the Australia courts: what's next
The release of the Federal Court's Practice Note raises a clear question about how other courts and tribunals will respond to the ever-changing use of AI in Australia, and its ongoing effect on confidentiality and privilege.
It remains to be seen whether similar or updated guidance will follow from other courts and tribunals. We may even see courts go further and formally implement the Victorian Law Reform Commission’s recommendations.
What is clear is that the Federal Court has taken a step in shaping practice expectations in this space. We expect AI, including where confidential material is concerned, will continue to emerge as a key feature of modern litigation.
Watch this space closely for further developments across Australia.
Should you wish to learn more, including about the intersection of AI and legal professional privilege, see our podcast series, ‘AI and legal professional privilege’ and article, ‘Beware of artificial intelligence and the potential waiver of legal professional privilege’ on this topic.
This article was prepared with assistance from Marcus Jones, Law Graduate.
Contact
