Source: Courts Administration Authority of South Australia

Courts Administration Authority of South Australia has published a formal set of guidelines regarding the use of generative artificial intelligence in litigation within South Australian courts. Issued on December 24, 2025, these guidelines respond to the increasing adoption of large language models by legal practitioners for drafting submissions and conducting research. The core principle of the document is that legal professionals remain personally responsible for the accuracy and integrity of any materials filed in court, regardless of whether those materials were assisted by AI.

The guidelines specifically warn against the danger of "hallucinations," where AI systems generate fictitious case law or legal citations. To mitigate this, lawyers are now strictly required to disclose when generative AI has been used in the preparation of a submission and must certify that every cited authority has been manually verified for accuracy. Furthermore, the court emphasizes the importance of maintaining client confidentiality, cautioning practitioners against inputting sensitive or privileged information into public AI models that may use such data for future training.

In addition to lawyer conduct, the guidelines address the use of AI-generated evidence and the potential for bias in automated systems. Judges in South Australia now have the authority to request detailed information about the AI tools used in a case if there is a suspicion of inaccuracy or ethical breach. This proactive judicial stance aims to preserve the public's trust in the legal system by ensuring that the introduction of advanced technology does not undermine the foundational principles of justice, accountability, and professional diligence.