GenAI, Fake Law & Fallout
CFLP report shapes judicial discussion as GenAI misuse escalates across courts
Published on the 28 November 2025
A major report from the UNSW Centre for the Future of the Legal Profession has detailed the growing misuse of generative AI in legal matters across ten jurisdictions, with its findings now noted by the Chief Justice of New South Wales in a public update on the review of Supreme Court Practice Note Gen 23.
The Centre for the Future of the Legal Profession at UNSW Law and Justice has released GenAI, Fake Law and Fallout, a significant research report examining 520 documented instances of generative AI misuse in legal proceedings between January 2023 and September 2025. Drawing from ten common law and hybrid jurisdictions, the report provides the most comprehensive cross country analysis currently available.
Australia recorded 87 cases during the study period, placing it second only to the United States, which reported 319.
Notable case numbers also arose in the United Kingdom, Canada, India and other jurisdictions surveyed. The data shows that the spread of inaccurate or fabricated AI generated material is a rapidly intensifying problem across courts and tribunals.
The importance of the research has been recognised at the highest judicial level. In a public letter issued in connection with the ongoing review of Supreme Court Practice Note Gen 23 on the Use of Generative AI, the Honourable A. S. Bell, Chief Justice of New South Wales, noted recent literature on AI in the courts and expressly referenced the report published by the Centre. The Chief Justice confirmed that the Practice Note, which took effect in February 2025 following extensive consultation, will continue to be reviewed due to the fast changing nature of generative AI.
The report identifies clear patterns in Australia and overseas.
Self-represented litigants make up more than 78 per cent of identified cases locally, with similar dominance seen in the United States, the United Kingdom and Canada.
Misuse typically arises where litigants rely on AI-generated submissions or citations that contain errors, omissions or invented authorities.
The research also shows that cases tend to cluster in lower courts, tribunals and commissions where self-representation is common. The number of detected cases rose sharply during 2025, indicating that concerns relating to generative AI are growing rather than settling.
GenAI, Fake Law and Fallout includes detailed visual analysis that charts jurisdictional comparisons, profiles of users and the movement of cases across the court hierarchy. It also considers the implications for professional responsibility, access to justice and community confidence in the administration of justice.
The project was funded by the Australian Academy of Law. Research was led by Vicki McNamara for the Centre for the Future of the Legal Profession at UNSW Law and Justice. The findings provide an essential evidence base for courts, practitioners and policy makers as they continue to assess the challenges and opportunities of generative AI in the legal system.
The CFLP academics and others will be discussing the report in detail at the 2026 Annual Summit, Truth, Trust & Technology: Implications of AI for Law. Click here for more information.