Melbourne Lawyer Investigated for AI-Generated Fake Citations in Family Court
The Melbourne division of the Federal Circuit and Family Court of Australia is housed in the Commonwealth Law Courts Building on the corner of La Trobe Street and William Street
Image Source: Adz, Public domain, via Wikimedia Commons
In a troubling development for the legal community, a Melbourne-based lawyer has come under investigation by the Victorian legal complaints body following the misuse of artificial intelligence software. The incident, which unfolded in a family court case, involved the submission of fabricated case citations generated by AI, leading to the adjournment of a critical hearing.
The Incident
During a hearing on July 19, 2024, an unnamed solicitor representing a husband in a marital dispute presented a list of prior cases as requested by Justice Amanda Humphreys. However, Justice Humphreys later revealed that neither she nor her team could verify the authenticity of the cases provided. Upon returning to court, the lawyer admitted that the list was created using Leap, a legal AI software, and acknowledged the involvement of AI in generating the citations.
Legal Proceedings
The solicitor conceded that he failed to verify the accuracy of the AI-generated information before submitting it to the court. Initially, he was given a month to justify why he should not be referred to the Legal Services Board and Commissioner for further investigation. In an August ruling, Justice Humphreys proceeded with the referral, emphasizing the importance of investigating professional conduct, especially as the use of AI tools in law becomes more prevalent.
In her judgment, Justice Humphreys highlighted that the solicitor had not reviewed the AI-generated list, resulting in the submission of nonexistent cases. Although the lawyer issued an unconditional apology and pledged to learn from the mistake, the court deemed the referral necessary to address the broader implications for legal practice.
Response from AI Provider
Leap, the AI software in question, responded by underscoring the critical role of verification in the legal profession. A spokesperson explained that Leap offers a free verification process where experienced local lawyers review the software’s output to ensure its accuracy. In this particular case, the verification was completed, but the solicitor did not utilize the corrected information provided by Leap, leading to the submission of false data.
The company emphasized that while AI tools like Leap are powerful and widely used—66,000 legal professionals globally rely on their software—it is imperative for users to apply them correctly to enhance legal practice rather than compromise it.
Broader Implications
This incident serves as a stark reminder of the potential pitfalls associated with the integration of AI in legal settings. As AI tools become increasingly sophisticated, the responsibility on legal professionals to understand their functionalities and limitations grows. Justice Humphreys pointed out that, although the family court has yet to establish specific guidelines for AI usage, existing guidelines from the Supreme Court of Victoria and the County Court of Victoria mandate that practitioners must be well-versed in how AI tools operate and their constraints.
Previous Cases
The Melbourne case is not isolated. Earlier incidents have highlighted similar issues with AI in the legal field. In March, a Canadian lawyer faced criticism after using ChatGPT to locate case law in a custody battle, only to discover that the AI had generated fictitious cases. Additionally, last year, lawyers representing a client injured on a flight encountered "bogus opinions" from ChatGPT when seeking supportive case law, leading to judicial disapproval.
Source: The Guardian
