Administrative Review Tribunal adopts practice direction on generative AI use in proceedings

Regulatory action may follow if legal representative breaches professional obligation, tribunal says

Administrative Review Tribunal adopts practice direction on generative AI use in proceedings

The Administrative Review Tribunal has released a practice direction on the use of generative artificial intelligence (genAI) in its proceedings, accompanied by an optional checklist to help tribunal users navigate their obligations under the new practice direction. 

According to the tribunal’s update, the practice direction aims to provide guidance on: 

  • the tribunal’s expectations of everyone involved in its proceedings, specifically in connection with genAI 
  • the tribunal’s potential considerations when giving directions and making orders about genAI use in a particular proceeding 
  • instances when the tribunal might require someone to verify, explain, or disclose genAI use 

The tribunal’s update explained that its new practice direction complements its 2026 practice directions on common procedures and expert evidence. 

Emilios Kyrou, the tribunal’s president, signed the new practice direction on 20 August 2026. The tribunal adapted its practice direction from the Federal Court’s practice note on the use of genAI (GPN-AI), published on 16 April 2026

Kyrou became a judge of the Federal Court of Australia in June 2023, an appeal judge of the Victorian Supreme Court’s Court of Appeal in July 2014, and a trial judge of the Victorian Supreme Court in May 2008. 

Administrative Review Tribunal’s expectations for genAI use

The new practice direction lists the tribunal’s expectations for genAI use relating to its proceedings. Specifically, the tribunal expects those utilising AI to: 

  • Have a basic understanding of the tool’s capabilities, limitations, and risks 
  • Ensure that genAI use does not adversely affect the administration of justice 
  • Comply with their legal and professional responsibilities 
  • Refrain from utilising the tool where inappropriate 
  • Disclose AI use, if required 
  • Provide certification, if requested 
  • Verify genAI-assisted material 
  • Explain how they used the tool and how they checked the material, if asked 
  • Refrain from relying solely on genAI to verify that tool’s own output 

Consequences of breaching genAI practice direction

According to the tribunal’s new practice direction, someone using genAI can face consequences if they use the tool inconsistently with the new practice direction or with a tribunal order or direction, depending on the nature, seriousness, and effects of the noncompliance. 

In the event of noncompliance, the tribunal can: 

  • Order the correction, replacement, or withdrawal of the affected material 
  • Ask for an explanation regarding how someone used genAI or checked the output 
  • Require additional evidence or submissions 
  • Direct attendance at a hearing 
  • Give less or no weight to the impacted material if noncompliance affected its reliability 
  • Refuse to rely on the output for lack of proof of its accuracy, source, or reliability 

“If the conduct of a legal or other professional representative may involve a breach of a professional obligation, the Tribunal may take any step available to it, including referring the matter to an appropriate regulatory body,” the tribunal noted in its practice direction

The tribunal clarified that it would usually consider such regulatory action against the representative, distinct from any procedural or evidentiary consequences for the party or their material, after issuing a substantive decision to ensure that the representative’s conduct would not negatively impact the party’s case. 

More court updates on artificial intelligence

Here are some other recent news stories from Australian courts that touch upon AI use. 

Last June, the County Court of Victoria released a practice note for court users and guidelines for judicial officers, stating the court’s expectations for the responsible AI use in court proceedings. 

In late May, William Alstergren, chief justice of the Federal Circuit and Family Court of Australia (FCFCOA), released a practice direction on AI use, developed by the FCFCOA’s AI committee and reflecting insights from internal and external consultation.

In mid-May, the Supreme Court of Victoria published a new practice note for court users and judicial guidelines for judicial officers on AI use in an effort to lead, support, and clarify the appropriate and innovative utilisation of this emerging technology. 

Enjoy this story? Read the latest litigation and dispute resolution news on the main page!