The Supreme Court of Victoria has released updated practice notes and judicial guidelines that outline acceptable uses of artificial intelligence in court proceedings while cautioning against excessive dependence on AI-generated materials without verification. These directives were published following an examination by the Victorian Law Reform Commission and expanding on earlier AI-related policies from 2024. With these measures, Victoria becomes one of Australia’s first legal systems to formalize such rules.

The practice note directed at court users permits the use of AI tools to help draft legal documents, but it explicitly states that legal professionals remain solely accountable for ensuring their filings are accurate and complete. Four primary risks are identified: the inclusion of irrelevant or incomplete information, AI-generated inaccuracies, privacy issues arising from external tool usage, and potential cost penalties for submissions containing errors. The court warns that relying on unverified AI content in legal arguments could lead to disciplinary action.

Judges receive separate guidelines that impose strict limitations. Judicial officers are prohibited from using generative AI to reach decisions or replace their own reasoning. Instead, AI may support administrative functions, such as locating relevant case law, summarizing evidence, or reviewing drafts, so long as it does not replace direct engagement with submissions or evidence. Chief Justice Richard Niall reinforced this stance, stating that AI should serve as an aid, not a substitute for judicial judgment: “Any use of AI will be an aid to, not a replacement of, judicial decision-making.”

The Judicial College of Victoria will conduct an online AI skills program exclusively for judicial officers to help judicial officers utilise AI consistently with judicial values and ethics, adapt AI systems to their needs, create queries that maximise output value, and examine the methods for gathering and testing answers.