Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
Firm news
08 June 2026
A recent decision by the New Zealand Supreme Court in Jones v Family Court at Whangārei [2026] NZSC 1 serves as a timely reminder about the limitations and risks attached to the use of artificial intelligence (AI) in legal proceedings.
In Jones, the applicant filed submissions for a leave application that included citations to numerous cases entirely hallucinated by AI. Several citations combined real case names with incorrect references; four genuine cases were misused to support propositions they did not stand for.
Jones involved a self-represented litigant, but the implications of incorrect use of AI are considerably more severe for lawyers (and their clients). While Jones made it clear that the unsupervised use of AI in court submissions can amount to an obstruction of justice or contempt of court, this is far from a uniquely New Zealand problem. Courts in the United States, England and Wales, and Australia have confronted the same phenomenon, responding with financial sanctions, formal practise notes, and regulatory referrals. The emerging picture across all jurisdictions is consistent: AI demands human oversight, not human abdication.
New Zealand already has formal guidance in this space. In December 2023, the Courts of New Zealand issued three sets of Guidelines for Use of Generative Artificial Intelligence in Courts and Tribunals (GenAI Guidelines), one each for lawyers, judges and judicial officers, and non-lawyers respectively. Several key themes run through these Guidelines:
GenAI chatbots may make up fictitious cases, citations and quotes that appear to come from real sources, and confirm that information is accurate if asked, even when it is not.
Anything entered into a public GenAI chatbot may be retained by the platform and used to respond to queries from other users. Lawyers must not enter private, confidential, suppressed or legally privileged information into any public AI tool.
Proactive disclosure of GenAI is required if a court or tribunal asks for it, but is not required as a matter of course.
GenAI should be used with caution for legal analysis, as it cannot critically examine patterns, does not produce a neutral output, and has limited knowledge of New Zealand law and procedure.
The Supreme Court in Jones endorsed the GenAI Guidelines, reiterating the requirement for all parties to verify the accuracy of AI-generated material before relying on it in proceedings:
“You are responsible for ensuring that all information you provide to the court/tribunal is accurate. You must check the accuracy of any information you get from a [generative AI] chatbot before using that information in court/tribunal proceedings.”
Jones is not the only New Zealand decision to have grappled with AI in proceedings. For example, in LMN v STC (No 2) [2025] NZEmpC 46, a self-represented litigant cited a supposed legal precedent to suggest that the Court will take financial constraints into account when making procedural rulings. The Employment Court found that no such case existed, and reiterated that information produced by AI must be independently verified before being relied upon in legal proceedings. The Court again referred to the GenAI Guidelines in support of its position. Further, in O’Driscoll v Rehab Co Mobile Limited [2026] NZERA 106, the Employment Relations Authority awarded increased costs against a self-represented party, in part because of the applicant’s reliance on cases that did not exist or could not be found. The Authority noted that, despite attention being drawn to the GenAI Guidelines, the applicant continued to refer to cases that did not exist in final submissions.
The problem of AI hallucinations in court filings has generated significant jurisprudence and formal regulatory responses across multiple common law jurisdictions.
In the United States, for example, an early example of such issues arose in 2023 in Mata v Avianca Inc. In that case, attorneys submitted ChatGPT-fabricated citations in federal court proceedings, and continued to maintain those citations when their existence was challenged. They were ultimately sanctioned US$5,000 and found to have acted in subjective bad faith. The court observed that using a reliable AI tool is not inherently improper, but that existing professional rules impose a “gatekeeping role on attorneys to ensure the accuracy of their filings”. In a separate case, in late 2025, a judge found that the plaintiff in that case had submitted AI-generated audio and video of a real person as authentic testimony. Whilst the video in that case seemed obvious, the increasing quality of deepfake evidence is a significant concern. More recently, in April 2026, the renowned Wall Street firm Sullivan & Cromwell formally apologised to the United States Bankruptcy Court for the Southern District of New York after it emerged that an emergency motion it had filed contained more than 40 inaccuracies, including fabricated case citations, misquoted provisions of the Bankruptcy Code, and non-existent legal sources generated by AI.
In England and Wales, in 2025 the English Divisional Court addressed the same phenomenon in R (Ayinde) v London Borough of Haringey, holding that freely available generative AI tools “are not capable of conducting reliable legal research”, and that lawyers have a professional duty to check AI-generated research against authoritative sources before using it. The court noted that the problem raises broader concerns about training, supervision, and regulation across the profession. More recently, in UK v Secretary of State for the Home Department, the Upper Tribunal Immigration and Asylum Chamber went further, confirming that supervisors who fail to ensure fee-earners’ AI-assisted work is accurate are likely to be referred to the Solicitors Regulation Authority, and warning that uploading confidential materials to a public AI tool such as ChatGPT constitutes a breach of client confidentiality and waives legal professional privilege.
In Australia, following the identification of incorrect use of legal cases by a legal practitioner, the Victorian Legal Services Board varied that lawyer’s practising certificate so that he was no longer entitled to practise as a principal solicitor and would only practise as an employee solicitor. The Federal Court of Australia’s recent guidelines have specific disclosure requirements in relation to the preparation of expert evidence. The New South Wales Supreme Court goes further, requiring that where AI has been used to prepare written submissions, the author must personally verify in the body of the document that all citations, legal and academic authority and case law and legislative references exist, are accurate, and are relevant. NSW also prohibits the use of GenAI to draft or prepare the content of an expert report (or any part of an expert report) without prior leave of the Court.
Careful use of AI certainly has its benefits. In Ale v Easthope, a self-represented defendant used ChatGPT to help prepare written submissions in a property possession proceeding. Associate Judge Gellert commended the result, noting that it produced submissions that were “well structured and easier to comprehend than might have otherwise been the case with a litigant in person”, and that the defendant's upfront disclosure around use of AI allowed her to assess the submissions accordingly.
In VA v JL, AI was used in translation and document analysis for foreign language materials. The Legal Complaints Review Officer found it more likely that messages had been deleted than that the AI-assisted log had been fabricated, and accepted ChatGPT-generated translations of a chat application log as the more complete version of the relevant messages.
AI can genuinely assist in litigation, including structuring submissions, translating documents, organising materials, but only where the human using it maintains meaningful control and oversight. All information generated by a GenAI chatbot must be independently checked before being used in proceedings; for lawyers, that check must be carried out by an appropriately qualified person.
Whether a lay litigant or a lawyer, anyone using AI in litigation should:
Independently verify every citation and legal proposition against authoritative sources, rather than asking the AI tool about its accuracy.
Avoid using AI uncritically for legal analysis. For lawyers, be aware that existing professional obligations, including the duty not to mislead the court, apply in full to any AI-assisted work.
Protect confidentiality by not entering any private, confidential, suppressed or legally privileged information into a public AI tool, and disclose AI use when required.
Disclose AI use transparently, as this supports rather than undermines credibility.
Be alert to authenticity issues and the potential for deepfake evidence, and ensure that any potentially incorrect evidence is challenged.
This article was co-authored by Raksha Tiwari (Solicitor), in our Litigation team.
Speak with one of our experts.
Read more of our related insights.
Legal update
03 June 2026
Podcast
27 May 2026
Podcast
30 April 2026
Podcast
17 April 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Publications and reports
17 February 2026
Legal update
26 September 2025
Podcast
14 August 2025
Publications and reports
05 August 2025
Legal update
05 August 2025
Legal update
31 July 2025
Firm news
28 July 2025
Podcast
26 June 2025
Legal update
05 May 2025
Podcast
29 April 2025
Podcast
29 April 2025
Publications and reports
09 April 2025
Podcast
20 February 2025
Publications and reports
12 February 2025
Publications and reports
12 February 2025
Legal update
15 November 2024
Podcast
06 November 2024
Publications and reports
31 October 2024
Publications and reports
31 October 2024
Publications and reports
31 October 2024
Legal update
24 October 2024
Podcast
16 October 2024
Legal update
08 October 2024
Legal update
09 September 2024
Legal update
06 September 2024
Toolkits and guides
27 August 2024
Legal update
11 July 2024
Legal update
09 July 2024
Legal update
09 May 2024
Legal update
29 April 2024
Legal update
17 April 2024
Legal update
27 March 2024
Legal update
26 March 2024
Legal update
14 March 2024
Legal update
11 March 2024
Podcast
28 February 2024
Podcast
28 February 2024
Podcast
28 February 2024
Podcast
21 February 2024
Podcast
21 February 2024
Publications and reports
15 February 2024
Publications and reports
15 February 2024
Publications and reports
15 February 2024
Podcast
15 February 2024
Publications and reports
15 February 2024
Publications and reports
15 February 2024
Publications and reports
15 February 2024
Podcast
15 February 2024
Firm news
15 February 2024
Publications and reports
12 February 2024
Podcast
12 February 2024
Legal update
08 February 2024
Publications and reports
08 February 2024
Firm news
08 February 2024
Legal update
25 January 2024
Firm news
17 January 2024
Opinion
16 January 2024
Publications and reports
30 November 2023
Publications and reports
30 November 2023
Opinion
14 November 2023
Legal update
01 November 2023
Legal update
18 October 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Publications and reports
31 August 2023
Legal update
24 August 2023
Legal update
08 August 2023
Legal update
07 August 2023
Publications and reports
28 July 2023
Legal update
12 July 2023
Case study
21 June 2023
Opinion
02 June 2023
Legal update
31 May 2023
Legal update
27 April 2023
Publications and reports
13 April 2023
Publications and reports
13 April 2023
Publications and reports
13 April 2023
Publications and reports
13 April 2023
Publications and reports
13 April 2023
Publications and reports
13 April 2023
Case study
03 March 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Publications and reports
09 February 2023
Firm news
07 February 2023
Publications and reports
04 November 2022
Publications and reports
04 November 2022
Publications and reports
04 November 2022
Legal update
19 October 2022
Firm news
14 July 2022
Legal update
05 April 2022
Legal update
04 November 2021
Legal update
20 May 2021
Legal update
10 February 2021
Legal update
09 February 2021
Legal update
24 July 2020
Legal update
05 March 2020
Legal update
06 December 2019
Legal update
02 December 2019
Opinion
24 October 2019
Opinion
17 July 2019
Legal update
15 November 2018
Sign-up to receive relevant insights and invitations from MinterEllisonRuddWatts.
Services
Auckland
Tāmaki Makaurau
PwC Tower
15 Customs Street West
Auckland 1010
+64 9 353 9700
Wellington
Te Whanganui-a-Tara
BNZ Place
1 Whitmore Street
Wellington 6011
+64 4 498 5000
© 2026 MinterEllisonRuddWatts