The Art of Intelligent Use of Artificial Intelligence: A Practical Guide for Native Title Lawyers

Key takeaways[1]

  • Courts have sanctioned lawyers for improper or unverified use of generative AI, including in a native title matter.
  • Generative AI can be a useful tool but must not replace diligence, professional judgment or supervision.
  • Lawyers must understand how Generative AI works, verify outputs, and ensure compliance with duties of candour and confidentiality.
  • Native title work presents unique challenge. For example, Generative AI lacks cultural sensitivity and understanding of First Nations contexts.
  • While Generative AI may assist with transcription, research, or summarising materials, human native title legal expertise remains essential for legal analysis.
  • Generative AI and Large Language Model use by, for example, unrepresented parties could increase risks to relationships and accuracy in native title matters.
  • As Generative AI evolves, native title practitioners must be proactive in addressing its use and implications with clients.

Introduction

Artificial Intelligence (AI) is everywhere. For lawyers, the role of generative AI (GenAI) including Large Language Models (LLM) is becoming increasingly prevalent, pitched as a transformative drafting and research tool.[2] There are noteworthy attempts by the legal profession to help lawyers navigate the rapidly evolving challenges of GenAI. Internationally,[3] this includes the “AI Task Force” of the American Bar Association publishing its first report in 2024,[4] followed by the UK Judiciary publishing its AI “Guidance for Judicial Office Holders” on 31 October 2025.[5] In Australia, it is anticipated the Federal Court of Australia will shortly publish a practice note in relation to AI,[6] with the Supreme Court of New South Wales issuing a practice note in January 2025.[7] Further, the College of Law’s Centre for Legal Innovation has produced a range of free resources for lawyers.[8] Interestingly, the Australian Guide to Legal Citation[9]remains silent on how to cite the use of AI.

This article explains the current state of play for lawyers navigating AI, highlighting key guidance and practice notes and recent cases. The article then explores the ethical and practical challenge of AI use in the native title context.

Current state of play

From a broader public confidence perspective, there is a growing public interest in the regulation of AI in professional services in Australia,[10] including in relation to the legal profession.[11] 

Regulatory concern about the risks of GenAI for the legal profession and the tension with professional and ethical obligation is justified. From false case citations to poor supervision, examples of GenAI undermining the reputation of lawyers are becoming more frequent.

There are five noteworthy and recent examples of improper use of GenAI in Commonwealth Courts,[12] including in relation to a native title claim.

  • Murray obh of Wamba Wemba Native Title Claim Group v State of Victoria [2025] FCA 731 (22 April 2025) (Murray):

This was the first time the use of AI in native title claims has been considered by the Court.[13] In Murray, Justice Murphy awarded costs against the then solicitor for the Applicant personally. This was because a junior solicitor prepared footnotes to an Amended Native Title Determination Application (a Form 1), using Google Scholar. The footnotes were then used in a summary document. The footnotes, however, were fabricated.[14] Justice Murphy found that, in addition to the judgment itself criticising the conduct, indemnity costs should be awarded against the solicitors personally because of the cost and delay to the parties and the compromise to the administration of justice.[15]

  • JNE24 v Minister for Immigration and Citizenship [2025] FedCFamC2G 1314 (15 August 2025) (JNE24):

In JNE24, Judge Gerrard ordered, but did not name,[16] a lawyer pay costs personally and referred conduct to the regulatory body for consideration. This concerned improper conduct by the lawyer of citing four cases in written submissions said to be in support of various propositions. These submissions relied on research through Claude AI.[17] The cases did not exist or did not stand as authority for the proposition they were said to support.[18] The Court declined to reproduce citations to reduce the risks of contributing to case hallucinations.[19] In considering whether to refer the lawyer to the regulatory body, Judge Gerrard commented that:[20]

What is expected from legal practitioners as part of their duty to the Court and to their client is that those cases (if they do exist) are reviewed to ensure they are authority for the principle the lawyer wishes to rely upon, have not been subsequently overturned or distinguished by a higher court, and are considered in respect of how and why those principles are relevant to the factual matrix of the case in which they intend to advance that proposition. Legal principles are not simply slogans which can be affixed to submissions without context or analysis.

  • Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (31 January 2025) (Valu):

In Valu, Judge Skaros referred, but did not name,[21] a lawyer’s conduct to the regulatory body for consideration. The relevant conduct included false citations and quotes in submissions, generated using ChatGPT,[22] and, correspondence with the Court without the consent of the other parties.[23] In assessing the lawyers conduct, Judge Skaros observed that:[24]

The conduct of the ALR, in filing an application and submissions which contained citations to Federal Court of Australia cases which do not exist and alleged quotes from the Tribunal’s decision which do not exist, falls short of the standard of competence and diligence that the applicant in the substantive proceedings was entitled to expect from his legal representative. The conduct also falls short of a legal practitioner’s duty to the Court, including the duty to ensure that the Court is not deceived or mislead, even if unintentionally: r 19.1 of the Conduct Rules.

  • Dayal [2024] FedCFamC2F 1166 (27 August 2024) (Dayal)

In Dayal, Judge A Humphreys referred Mr Dayal’s conduct to the regulatory body for consideration. The relevant conduct was the tendering of a list and summary of authorities which did not exist, using an AI research tool that was incorporated in the legal practice management software LEAP.[25] Mr Dayal did not verify the accuracy of the information.[26]

  • JML Rose Pty Ltd v Jorgensen (No 3) [2025] FCA 976 (19 August 2025) (JML)

In JML, an unrepresented litigant used AI to generate submissions with fake authorities.[27] In commenting generally on the risks of AI in legal proceedings, Justice Wheatley observed (citations omitted):[28]

The use of generative AI to prepare submissions that may include fake authorities will nearly always introduce added costs, complexity and add to the burden of other parties and to the Court: Costaras at [16] and [49]. I gratefully adopt the observation from Ayinde at [9] that “(t)here are serious implications for the administration of justice and public confidence in the justice system if artificial intelligence is misused.” The Court in Ayinde observed from [10] to [31] the existing guidance, regulatory duties of the profession, referrals and the Court’s powers. Matters which are within the Court’s own domain include in the most serious of cases contempt of Court. The observations in Ayinde at [26]–[28] regarding contempt of court are not limited to legal practitioners.

Several legal profession bodies have issued statements and guidance notes about the use of GenAI,[29] with one body issuing a template policy for legal practices on the responsible use of GenAI.[30] The Law Council of Australia has also published an article about the limitations on the use of AI.[31] Ten key aspects of these articles and guidance notes are as follows:

  1. Lawyers’ paramount duty is to the Court and to act in the interests of justice.
  2. The use of GenAI is subject to duties of candour to the Court, diligence and confidentiality.
  3. Lawyers must not mislead and should disclose assistance provided by AI programs.
  4. Lawyers must understand how GenAI tools work and their limitations, including currency, risks of inaccuracies, bias and incomplete searches.
  5. Lawyers must verify all GenAI outputs.
  6. Lawyers must exercise professional judgment and skill, and expertise in the relevant area is required.
  7. Supervising lawyers and law practices must ensure proper training and supervision in the use of GenAI.
  8. Legal practices ought to develop and maintain a GenAI policy and carefully chose service provides, if it is to be used regularly, with up-to-date precedents (like terms of engagement) that sufficiently address any use of GenAI.
  9. Lawyers using GenAI in legal services need to ensure costs are still charged fairly.
  10. Lawyers should consider how a potential new or existing client’s use of GenAI ought to be considered and addressed.

What does this all mean for lawyers navigating the ethics of GenAI?

As the various guidance and practice notes are anticipating, the practise of law is rapidly changing because of GenAI. The ethical issues that arise through lawyers’ use of GenAI is no longer theoretical. The issues are real.

Thomson Reuters’ ‘2025 Future of Professionals Report’, surveyed legal professionals about their AI use. The survey results indicated that lawyers are primarily using AI for document review, legal research, summarising documents and drafting briefs or memos.[32] It goes without saying that lawyers must always act professionally and ethically, and in the interests of justice. Lawyers who properly use GenAI as a tool must accept that it can only be a step in the lawyering process.

In relation to using GenAI in legal research, lawyers ought to remind themselves of an express duty to exercise all due care and skill when giving advice to and acting for a client. This means that lawyers have positive obligations to their clients to ensure that they have exercised their care and skill in preparing and delivering advice. This includes by carefully reviewing citations and ensuring that cases relied upon in support of propositions support the propositions advanced.

In relation to using GenAI in drafting and correspondence, lawyers ought to remind themselves of the need to promote public confidence in the justice system and for court efficiency.[33] Correspondence that has been drafted using GenAI, usually filled with unnecessary emphasis like em dashes, is relatively obvious and will likely lead to embarrassment and impacts to the reputation of the lawyer. Further, the time it takes to identify and correct GenAI’s errors ought to cause lawyers to reflect on perceived drafting efficiency. However, GenAI will get better. As a result, it will get harder to spot GenAI’s influence and errors, even for lawyers trained to understand and unpick. At the end of the day, lawyers cannot rely on GenAI to communicate effectively or supervise.

In other words, lawyers who rely on GenAI to replace lawyering will likely be in serious breach of their ethical and professional obligations.

GenAI in the native title context

Native title is complex for several reasons. Not only is the area of the law exceptionally complex in and of itself, but the operating environment is also challenging. For the already overworked minds of hardworking native title lawyers, GenAI adds to the complexity.

Consider, for example, the following in the context of native title.

  • GenAI may, generally, be a useful transcription tool in some settings. In addition to the obvious confidentiality obligations, and the concerns that LLMs uses data it collects from confidential information as part of its language processing, GenAI does not (currently) understand First Nations languages or place names, or, pick up on communication nuances within a community. Those native title lawyers who have been involved in, for example, court transcript corrections will appreciate the complexity of attempting to transcribe witness evidence in a native title trial.
  • GenAI may, generally, be a useful tool to prepare summaries and assess large volumes of materials in some settings. In the native title context, GenAI may become a tool to review anthropological reports, genealogical reports and witness statements. However, these materials are inherently complex and sensitive. GenAI would be easily confused, for example, by the repetition of names in a family tree to interpret the information as if the repetition of names were the same person. A further example is the likely confusion by the slight difference in description and spelling of a language or place name, with GenAI processing the information as if it were different languages or places. More importantly, an anthropological report for example, would need to be evaluated in terms of its probity value, which would require legal analysis. A lawyer cannot discharge their duties if they are relying on GenAI for that type of analysis.            
  • GenAI may generally be a useful legal research tool. GenAI could, for example, be a step in the process of narrowing the field when starting legal research to avoid a fishing expedition. There is, however, the broad question of applicability of LLM processing (which is based on probability of linguistic sequencing) to tasks requiring evaluation of principles of particular facts. That is, all cases are fact specific, and native title lawyers will appreciate the complexity and uniqueness of each native title case. Consider, for example, the complex facts and legal context relevant to appropriately assessing native title compensation. There is a high risk that native title cases are easily misunderstood by GenAI to incorrectly be applied in another matter, or, misunderstood to support a native title proposition that it does not.
  • For unrepresented litigants, GenAI may be a useful tool that expands access to a large amount information that would have been otherwise unobtainable. However, GenAI and LLMs will use information inputs to create responses in other situations. In other words, GenAI and LLMs may influence the views and perceived truths for Aboriginal and Torres Strait Islanders seeking answers in a complex native title system. This is dangerous in the native title context, where preserving relationships is paramount.  

Quick and loose predictions

The following predictions are purposefully ‘quick’ and admittedly no better than ‘loose’. We don’t know what GenAI will mean for native title lawyers.

Still, some predictions follow.

  1. It may take a while for GenAI and LLMs to process native title concepts more accurately than what appears currently available, compared to other legal fields (for example, high volume commercial transactions). Once it does, it will be hard to spot the use of GenAI, and it will become more important than ever for native title lawyers to exercise good judgment and problem-solving skills to reality test GenAIs analysis. For example, native title lawyers will need to quickly but properly assess whether GenAI’s response makes any practical sense for their client in that matter, at that time.
  2. Once GenAI becomes more familiar with native title concepts, we may see an increase in lawyers attempting to practise in native title. This could include, for example, legal practices that are relying on litigation funders for native title compensation claims.
  3. First Nations’ clients and unrepresented parties will increasingly use and rely on GenAI to answer complex questions on, for example, family history, contests on rights to speak for county and traditional boundaries and entitlement and valuation of native title compensation. This is high risk. As a result, native title practitioners ought to start proactively addressing GenAI with clients now.
  4. GenAI will get better at identifying relevant native title cases or agreement examples. Equally, sound legal research and analysis, combined with getting the basics of legal citations right, will become more valuable for native title lawyers.  
  5. We may start to see GenAI produced materials being framed as authentic evidence in native title matters. That could include falsified images, genealogies, historical records or audio recordings. It will be increasingly important for native title lawyers and anthropologists alike to analyse the authenticity and source of documents.  
  6. GenAI’s impact on the way we communicate in writing will increase, with messaging more likely to be lost in GenAI translation. As a result, it will become more important for lawyers to continually improve communication skills. In the native title context, this will mean it will be even more important to have ‘on country’ face-to-face discussions in a culturally respectful and empathetic way, and, lawyers who are expert advocates or facilitators as required. Lawyers with those types of soft-skill experience will become even more important to the native title system. For more information about effective communication in court processes, see our additional resource here.

This article was prepared as part of MPS Law’s professional development series for Native Title Representative Bodies and Service Providers and is not legal advice. For more information, contact us.

[1] A version of this summary was prepared by ChatGPT on 6 November 2025, The summary was, however, proof-read and edited for accuracy.

[2] MPS Law has purposefully not referenced any specific legal GenAI products in this introduction.  

[3] See https://www.techieray.com/GlobalAIRegulationTracker for a resource that tracks international developments, although MPS Law does not verify the accuracy of the resource.

[4] https://www.americanbar.org/content/dam/aba/administrative/center-for-innovation/ai-task-force/2024ai-taskforce-report-1-31-25.pdf

[5] https://www.judiciary.uk/wp-content/uploads/2025/10/Artificial-Intelligence-AI-Guidance-for-Judicial-Office-Holders-2.pdf

[6] See Federal Court of Australia “Notice to the Profession: Artificial Intelligence Use in the Federal Court of Australia” media release (29 April 2025), available at https://www.fedcourt.gov.au/law-and-practice/practice-documents/notice-to-profession/29-april-2025

[7] See https://supremecourt.nsw.gov.au/documents/Practice-and-Procedure/Practice-Notes/general/current/PN_SC_Gen_23.pdf but see https://cartlandlaw.com/nsw-genai-practice-note/ for commentary on the practical difficulties with the practice note.

[8] See https://www.cli.collaw.com/upcoming-events/2023/06/07/ai-for-legal-series

[9] https://law.unimelb.edu.au/__data/assets/pdf_file/0005/3181325/AGLC4-with-Bookmarks-1.pdf

[10] https://www.theguardian.com/australia-news/2025/oct/06/deloitte-to-pay-money-back-to-albanese-government-after-using-ai-in-440000-report

[11] https://www.abc.net.au/news/2025-11-03/legal-services-board-of-victoria-warns-lawyers-about-ai-in-court/105916628

[12] For a comprehensive list of recent cases from other jurisdictions, see Law Library Victoria, “Artificial Intelligence (AI) in the Law and Hallucinations”, available at https://www.lawlibrary.vic.gov.au/resources/guides/artificial-intelligence-ai-law-and-hallucinations, accessed on 4 November 2025. For a recent example, in South Australia, see South Australian Employment Tribunal, “Referral to the LPCC for the citation of fake cases” (16 October 2025), available at https://www.saet.sa.gov.au/2025/10/16/referral-to-the-lpcc-for-the-citation-of-fake-cases/.

[13] See MPS Law, “Case alert: A warning on the use of Artificial Intelligence in native title claims – Wamba Wemba Native Title Claim [2025]”, available at https://www.mpslaw.com.au/news/case-alert-a-warning-on-the-use-of-artificial-intelligence-in-native-title-claims/, and, Native Title News — August 2025 (2025) 15(5) NTN 98.

[14] Murray obh of Wamba Wemba Native Title Claim Group v Victoria [2025] FCA 731 at [5].

[15] See, generally, Native Title Act 1993 (Cth) s 85A(2).

[16] JNE24 v Minister for Immigration and Citizenship [2025] FedCFamC2G 1314 (15 August 2025) at [35].

[17] Ibid at [14].

[18] Ibid at [26].

[19] Ibid. See, also Luck v Secretary, Services Australia [2025] FCAFC 26; BC202502886 at [14], for an example of the Full Federal Court redacting the citation of a case that did not exist and “may be a product of hallucination by a large language model…so that the false information is not propagated further by artificial intelligence systems having access to these reasons.”

[20] Ibid at [34].

[21] Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (31 January 2025) at [36].

[22] Ibid at [22].

[23] Ibid at [25]. See, generally, MPS Law, “10 Tips for Lawyers on Communicating During Federal Court Proceedings”, available at https://www.mpslaw.com.au/premium-resources/federal-court-communication-tips/.

[24] Valu v Minister for Immigration and Multicultural Affairs (No 2) [2025] FedCFamC2G 95 (31 January 2025) at [18].

[25] Dayal [2024] FedCFamC2F 1166 (27 August 2024)at [1] and Handa v Mallick [2024] FEDCFAMC2F 957; BC202412301 at [7].

[26] Dayal [2024] FedCFamC2F 1166 (27 August 2024)at [1] and [18].

[27] JML Rose Pty Ltd v Jorgensen (No 3) [2025] FCA 976 (19 August 2025) at [7], [98].

[28] Ibid at [104].

[29] See, for example, Law Society of New South Wales, Legal Practice Board of Western Australia and Victorian Legal Services Board + Commissioner, “Statement on the use of artificial intelligence in Australian legal practice”; New South Wales Bar Association, “Issues Arising from the Use of AI Language Models (including ChatGPT) in Legal Practice” (12 July 2023); Legal Practitioners’ Liability Committee, “Limitations and risks of using AI in legal practice” (17 August 2023), Queensland Law Society, “Guidance Statement: No.37 Artificial Intelligence in Legal Practice”, available at https://www.qls.com.au/practising-law-in-qld/ethics/guidance-statements/no-37-artificial-intelligence-in-legal-practice (accessed on 4 November 2025).

[30] Queensland Law Society, “QLS Template AI Use Policy”, available at https://www.qls.com.au/Content-Collections/Template/QLS-Artificial-Intelligence-Policy-template, accessed on 4 November 2025.

[31] Law Council of Australia, “Artificial Intelligence and the Legal Profession”, available at https://lawcouncil.au/policy-agenda/advancing-the-profession/artificial-intelligence-and-the-legal-profession (accessed on 4 November 2025).

[32] Thomson Reuters, Marjorie Richter J.D., ’How AI is transforming the legal profession’, https://legal.thomsonreuters.com/blog/how-ai-is-transforming-the-legal profession/#heading-1. 

[33] See, for example, Federal Court of Australia Act 1976 (Cth) ss 37M and 37N.