Dayal (2024): AI Utilization in Australian Justice System

by

 Efe Gunes

CyJurII Scholar

on 4 August 2026

Citation: Dayal [2024] FedCFamC2F 1166, with the file number of MLC 10532  of 2024

I. Introduction 

This case law commentary examines Dayal [2024], decided by Judge A. Humphreys on 27 August 2024 at the Federal Circuit and Family Court of  Australia (Division 2). The case concerns Mr. Dayal (to be referred to as ‘the solicitor’for the purposes of  privacy), a solicitor, who tendered to the court a list and summary of legal authorities that didn’t exist. The  solicitor later admitted that the list and summary were prepared by an artificial intelligence (AI) tool, and  that he didn’t verify the ‘accuracy of the information generated by the research tool before submitting it to  the court’. This case matters as it provides the fundamental perception on the use of AI in litigation in the  Australian justice system and the practical consequences that follow when practitioners fail to verify AI generated legal research. This case law commentary first provides the facts of the case, followed by the  legal issues and reasoning that the court followed, and concludes with how the reasoning of the court is persuasive and provides necessary and important implications for future litigants and AI law in Australia. 

II. Facts 

The central party in this case is Mr. Dayal, a Victorian solicitor and principal of a firm referred to as C Law  Firm. The proceeding arose out of the court’s earlier decision in Handa & Mallick [2024] FedCFamC2F 957, in which the solicitor had appeared as agent and tendered a list of case authorities that was later found to be inaccurate. Following this, on 19 July 2024, Judge A. Humphreys delivered reasons on how the content of the list and summary of authorities was identified to be inaccurate and made an order for the solicitor to submit reasons as to why he ought not be referred to the Office of the Victorian Legal Services Board and Commissioner. 

In compliance with the court’s deadline, the solicitor filed written submissions on 19 August 2024, in which he acknowledged that: 

- He handed up to the court on 19 July 2024 a document that purported to contain summaries of  relevant authorities and included what looked like medium neutral citations identifying those  decisions; 

- He used legal software and an AI driven research tool to generate the list of authorities and  summaries; 

- Neither he nor any other legal practitioner at his firm had reviewed or verified the output generated  by that tool; and 

- The authorities identified in the list and summary didn’t, in fact, exist. 

Furthermore, the solicitor offered an unconditional apology to the court for tendering the inaccurate list and provided an assurance that he will ‘take the lessons learned to heart and will not commit any such further breach of professional standards in the future.’6 He also asked for the judge to not make a referral to the Victorian Legal Services Board. 

On 27 August 2024, Judge A. Humphreys delivered the final order based on the apology and the  acknowledgements of the solicitor. 

III. Legal Issues 

The primary legal question the court had to answer was whether the solicitor should be referred to the Victorian Legal Services Board and Commissioner. The judge specifically argued that the regulatory body is the ‘independent statutory authority and officer in Victoria whose role and functions include ensuring the effective regulation of the legal profession and the maintenance of high ethical and professional standards.’ 

A possible secondary legal issue is whether the genuine apology offered by the solicitor, and his acknowledgements could mitigate the impact of his conduct on the other party, which was also considered by the judge.

IV. Holding and Reasoning 

The court accepted the solicitor’s apology, noting that it is ‘genuine’ given the ‘stress [the solicitor] had experienced as a result of the consequences flowing from his conduct’. Notwithstanding this, the court held that it is in the public interest for the regulatory body of the Legal Services Board to be aware of the professional conduct issues arising in this matter, given the increasing use of AI tools by legal practitioners in litigation more generally. As a result, the court referred the case to the Office of the Legal Services Board and Commissioner. 

This reasoning was based on 2 separate sections. First, the court addressed the use of AI in litigation directly,  observing that AI sits ‘at the frontier of technological advances in legal practice and the conduct of  litigation.’ While the court didn’t focus on this section on the impact of contemporary decisions on the  use of AI in litigation in future practitioners, it did cite the USA District Court case of Mata v Avianca Inc which considered the attention to the risk of relying on GenAI for research purposes without independent verification. In that case, the USA District Court outlined the potential harms flowing from the filing of bogus submissions including: 

- Wasting time and financial resources for the opposing party; 

- Wasting the time of the court; 

- Depriving the client’s arguments; 

- Damaging the reputation of judges and courts. 

The court reinforced this reasoning by reference to guidelines issued by the Supreme Court of Victoria and  the County Court of Victoria, which relevantly provide that: 

- Parties and practitioners who are using AI tools in the course of litigation should ensure they have  an understanding of the manner in which those tools work, as well as their limitations; - The use of AI programs shouldn’t mislead another participant in the litigation process as to the  nature of any work undertaken; 

- The use of AI must be subject to the obligations of legal practitioners.

On this basis, the court held that generative AI doesn’t relieve a practitioner of the need to exercise  independent judgment and professional skill in reviewing material before it is placed before the court. 

Second, the court considered the duties owed by legal practitioners more generally. Central to this reasoning was Legal Profession Uniform Law Application Act 2014 (Vic) s30, which sets out the objectives of the Victorian Legal Services Board – including ensuring the effective regulation of the legal profession, addressing concerns of the clients on law practices and legal practitioners, ensuring the adequate management of trust accounts and contributing to the effective administration of the Legal Profession Uniform Framework. 

In addition to this, the court considered other relevant duties of the legal practitioners including the paramount duty to the court and to the administration of justice, delivering legal services competently and diligently and not engaging in conduct which is likely to diminish public confidence in the administration of justice or bring the legal profession into disrepute. 

On these combined grounds and in light of the solicitor’s own acknowledgements – the court found that a breach of the professional standards expected of a solicitor in this court had occurred. 

V. Commentary 

Dayal [2024] can be considered a landmark case within the artificial intelligence (AI) law field in Australia, both for the court’s ultimate reasoning and for its sustained engagement with the use of AI in litigation and the standards expected of practitioners going forward. Judge Humphreys’reasoning is persuasive precisely because it strikes a measured balance: it neither prohibits nor discourages the use of AI tools by practitioners outright but instead reminds current and future litigants of the risks those tools carry – in particular, the risk of hallucinated or inaccurate citations – and of the professional consequences that follow when those risks aren’t properly managed, whether by their firm, their regulatory body, or the wider public. 

Before the delivery of the final order, the court delivered the perspective of the solicitor, including the fact  that he offered an unconditional apology, and how the circumstances led him to rely on AI without verifying  the generated information. At the end, the court did consider this perspective of the solicitor, and  acknowledged the genuine apology, before making the referral to the regulatory body. 

At the same time, the court was careful to ground the referral decision in principle rather than in the facts  of this solicitor’s conduct alone. Its explicit reliance on the Office of the Victorian Legal Services Board  and Commissioner as the appropriate body – informed by its earlier discussion of practitioners’ professional  duties – shows a reasoning process built on established regulatory structure rather than ad hoc disapproval. 

Moreover, the court considered the ‘increasing use of AI tools by legal practitioners in litigation’18 and that  public interest must be a factor in the referral to the regulatory body. It signals that the court viewed this  case as one of public interest rather than an isolated instance of individual carelessness. Therefore, the  reliability of AI assisted legal work bears not only on the accuracy of material placed before a court, but on  public confidence in the profession. 

Give that AI law remains a developing field, Dayal [2024] gave the court a substantial opportunity to  establish reasoning with broader prospective application, and Judge Humphreys’ approach delivers on that  opportunity. The reasoning cautions future litigants on verifying the accuracy of the generated content by  AI, and reminds them of their duties as practitioners, not only to the court, but also the administration of  justice, the clients and the general public.  

That said, the decision isn’t without limitations. It is worth noting that the Legal Profession Uniform Law has, to date, been adopted only in New South Wales, Victoria and Western Australia; other jurisdictions  regulate their legal professions under separate frameworks, and the court might usefully have acknowledged  this jurisdictional limitation more explicitly, given the decision’s likely influence beyond just Victoria.  Similarly, while the court’s discussion of AI-related risk is valuable, it remains tied closely to the facts of  the solicitor’s conduct rather than articulating a more general verification framework for AI-generated legal research. A dedicated section addressing that broader risk – appropriate, given the case’s likely status as a leading Australian authority on AI in litigation – would have strengthened the decision’s value as prospective guidance for the profession. 

VI. Conclusion 

Dayal [2024] stands as an early landmark Australian authority on the professional responsibilities expect of legal practitioners on the use of AI in litigation. Judge Humphreys’reasoning is persuasive in its balance as it neither bans the use of AI tools by practitioners nor treats their use as inherently improper but insists that any efficiency gained from AI must be matched by an ongoing duty of verification, diligence, candour  and competence. By reasoning through both the practical harms identified in Mata v Avianca and the specific duties owed by Victorian solicitors under the Legal Profession Uniform Law framework, the court grounded its referral decision in principle rather than in the severity of this solicitor’s conduct alone – a point reinforced by the court’s acceptance of his apology and its explicit statement that the referral wasn’t intended to be punitive. 

That said, as this commentary has noted, the decision leaves some questions unresolved. Because the Legal  Profession Uniform Law currently applies only in New South Wales, Victoria and Western Australia, the  guidance offered in Dayal [2024] doesn’t translate uniformly across Australian jurisdictions, and  practitioners regulated under other state or territory jurisdictions that may need to look elsewhere for  equivalent standards. The court also stopped short of articulating a general verification framework for AI  generated legal research, confining its reasoning instead to the facts before it and the existing guidelines of  the Supreme Court and County Court of Victoria. 

Given the increasing use of AI and its rapid development, it seems likely that Dayal [2024] will be cited as persuasive authority in future matters involving AI related misconduct, and that courts and regulators alike will be called upon to develop more comprehensive guidance in this area. In the meantime, the case stands  as a clear warning to practitioners: responsibility for the accuracy of material placed before a court can’t be delegated to an AI tool, however sophisticated, and a genuine apology – while capable of mitigating the consequences of a breach – doesn’t excuse the underlying failure to verify. 

VII. Bibliography 

A Articles 

Will Stidston, Barry Nilsson, ‘Artificial Intelligence in Legal Practice: Lessons from Dayal [2024]’, Thomson Reuters Practical Law (Article, 24 October 2024), < https://anzlaw.thomsonreuters.com/w-044- 7597?transitionType=Default&contextData=(sc.Default)&firstPage=true>  

B Cases 

Dayal [2024] FedCFamC2F 1166 < 

https://www.lexisnexis.com.au/ContentNotifier/URJNotifier/LinkMan?jurisdiction=fmca&bcnum=24143 79.htm>  

Handa & Mallick [2024] FedCFamC2F 957 

Mata v Avianca Inc. 678 F Supp 3d 443, 448 (SDNY, 2023) 

C Legislation 

Legal Profession Uniform Law Application Act 2014 (Vic)