A litigation file is rarely a single document. It is a claim or charge sheet, later pleadings, affidavits, correspondence, a bundle, a set of directions, and a string of dates that each depend on an event and a rule. Keeping all of it present, accurate and explainable — while also appearing in court — is ordinary litigation work, not an exception.

Tools that can read, sort, summarise and draft are already part of this picture. The American Bar Association has recorded that lawyers already use artificial intelligence to improve the efficiency and quality of legal services, including to review very large sets of documents in discovery.[1] England and Wales’s Divisional Court has observed that artificial intelligence “is a powerful technology” and “can be a useful tool in litigation, both civil and criminal”, including in large disclosure exercises.[2] Malaysia’s Bar has said that generative AI’s potential “to improve efficiency, manage time, reduce costs, and support creativity is undeniable.”[3]

The profession’s question is not whether AI has a place in legal work. It is how that work can be done without handing over verification, confidentiality, supervision and accountability.

Why AI has a place in legal work

Litigation produces paper, and more of it as a matter proceeds. Time spent hunting for a working affidavit, or reconstructing how a deadline was counted, is time a client pays for and a court may later ask about. Tools that scan large volumes or produce a first draft do not replace judgment; used carefully, they return time to the work only a lawyer can do. Circular No. 242/2025 lists high-volume review among the tasks where generative AI may, with caution, help identify themes or risk areas, and records benefits that include saving time and improving accessibility when drafting.[3] None of that is a licence to skip review.

There is also a public interest in consistency, provided a human remains responsible. In January 2020 the then Chief Judge of Sabah and Sarawak, Tan Sri David Wong Dak Wah, said the High Courts there had taken a first step by “getting the machine to determine the appropriate sentences for certain criminal offences,” that this would “only act as a guideline to the judicial officers,” and that, “When use appropriately, AI will increase efficiency of a process and raise the quality and consistency of our decisions.”[4] That is not a lawyer using a chatbot to draft a submission. It does show Malaysian courts treating AI as an assistant, while leaving the decision with the human.

Malaysia’s National Guidelines on AI Governance and Ethics, published by the Ministry of Science, Technology and Innovation in 2024, are voluntary for stakeholders. Their seven principles include reliability, safety and control; privacy and security; transparency; accountability; and the pursuit of human benefit and happiness.[5] The OECD’s intergovernmental AI principles, updated in 2024, likewise ask those who build and use AI systems to pursue beneficial outcomes, provide for human agency and oversight, and remain accountable for how the systems function.[6]

None of these sources treat AI as improper in itself. In Mata v Avianca, Inc., Judge P. Kevin Castel of the Southern District of New York wrote that there is “nothing inherently improper about using a reliable artificial intelligence tool for assistance,” but that existing rules impose “a gatekeeping role on attorneys to ensure the accuracy of their filings.”[7] The Singapore courts’ guide, applying from 1 October 2024, “does not prohibit the use of Generative AI tools to prepare Court Documents,” provided it is followed. The court user remains fully responsible, and lawyers must ensure that materials put before the courts are “independently verified, accurate, true, and appropriate.”[8]

Why the same work must be governed

The risk that has made headlines is the invented citation. In Mata, counsel submitted “non-existent judicial opinions with fake quotes and citations created by the artificial intelligence tool ChatGPT,” and then continued to stand by those opinions after the court questioned them.[7] In June 2025 the Divisional Court in R (Ayinde) v London Borough of Haringey, heard together with Al-Haroun v Qatar National Bank, considered cases referred because written arguments or statements produced with, or suspected to have been produced with, generative AI had not been checked, so that false information — “typically a fake citation or quotation” — was put before the court.[2]

Freely available generative artificial intelligence tools, trained on a large language model such as ChatGPT are not capable of conducting reliable legal research. … They may cite sources that do not exist. They may purport to quote passages from a genuine source that do not appear in that source.

Dame Victoria Sharp P., Ayinde [2025] EWHC 1383 (Admin) at [6]

Those who use such tools for research anyway have “a professional duty therefore to check the accuracy of such research by reference to authoritative sources” before advising a client or addressing a court — whether they used the tool themselves or relied on someone else who did, “no different from the responsibility of a lawyer who relies on the work of a trainee solicitor or a pupil barrister.”[2]

Malaysia’s Bar has said the same in domestic terms. Circular No. 242/2025, dated 3 July 2025, supplements Circular No. 342/2023 and reminds members that “all GenAI output must always be subject to human oversight, legal judgment, and ethical consideration.” Its risk table tells practitioners to verify output against official legal databases where citations may be hallucinated; to treat output “as suggestions only; never as authoritative”; to ensure that “all decision-making must be done by qualified legal professionals”; and to “maintain clear records of professional oversight and final review details for record and audit purposes.”[3]

Guidance for judicial office holders in England and Wales, refreshed on 31 October 2025, says information from an AI tool “must be checked before it is used or relied upon,” including cited sources that “might also be hallucinated.” Tools may “make up fictitious cases, citations or quotes.” Holders retain “personal responsibility for anything produced in their name.”[9] Singapore’s guide requires court users to verify that AI-generated references “actually exist and stand for the legal positions that are attributed to them,” and that extracts are quoted accurately.[8] Verification is not a courtesy. It is the difference between assistance and a filing the court cannot trust.

Confidentiality does not pause for a convenient tool

A lawyer’s duty of confidence is not suspended because a prompt box is open. Under rule 35(b) of the Legal Profession (Practice and Etiquette) Rules 1978, made under the Legal Profession Act 1976, an advocate and solicitor “shall preserve his client’s confidence and this duty outlasts his employment.”[10] Rule 16 requires the same advocate and solicitor, while acting with due courtesy to the tribunal, to “fearlessly uphold the interest of his client, the interest of justice and dignity of the profession.”[11]

Circular 242/2025 tells members not to use generative tools to process personal data, or to disclose information subject to solicitor-client confidentiality or legal professional privilege; to redact sensitive data before any input; and to avoid putting personal data or “client- or employee-specific facts into any GenAI tool, including those that are paid subscriptions.”[3] The United Kingdom judicial guidance is blunt: do not enter private or confidential information into a public AI chatbot; anything entered there “should be seen as being published to all the world.”[9]

ABA Formal Opinion 512 (29 July 2024) lists the same duties when lawyers use generative AI: competent representation, protection of client information, communication with clients, supervision of employees and agents, meritorious claims, candor toward the tribunal, and reasonable fees. Competence includes understanding the tool’s benefits and risks, remaining alert to “hallucinations,” and applying “an appropriate degree of independent verification or review.” The tools “cannot replace the judgment and experience necessary for lawyers to competently advise clients.”[1]

Supervision, and a record that can be explained

Governance is not only about catching a fake case. It is about knowing who did what. Ayinde treated reliance on an unchecked AI draft as no different, in principle, from reliance on unchecked work by a junior lawyer.[2] The Malaysian Bar asks firms to train staff, set internal policies, and keep records of oversight and final review.[3] The OECD principles call for human oversight and for traceability of processes and decisions.[6]

The EU Artificial Intelligence Act goes further for a different audience. Annex III classifies as high-risk AI “intended to be used by a judicial authority or on their behalf to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts.”[12] That does not, by itself, regulate a Malaysian law firm. It shows how seriously another major legal system treats AI that touches the administration of justice.

A governed practice is one in which a named person can explain a date, a draft or a citation; confidential instructions are not poured into a public chatbot; and a proposed step is distinguishable from an approved one.

Malaysian procedure still expects the lawyer, not the tool

None of this displaces the Rules of Court 2012. Order 1A requires the Court or a Judge, in administering the Rules, to have regard to “the overriding interest of justice and not only to the technical non-compliance with these Rules.”[13] Order 2 rule 1(2) adds that the Rules are “subject to the overriding objective of enabling the Court to deal with cases justly,” and that “the parties are required to assist the Court to achieve this overriding objective.”[13]

A file that is in order — dates that can be explained, papers that can be found, a bundle that matches its index — is one practical way a firm assists that objective. It is not a claim that software satisfies Order 1A. The duty remains the lawyer’s.

What this means in practice

Read together, the authorities point to a stable set of habits:

  • Use AI where volume or a first draft would otherwise consume time that should be spent on judgment — not as a substitute for a law report, a statute or a trained mind.[3]
  • Check every citation, quotation and proposition against an authoritative source before it reaches a client or a court.[2][3][8][9]
  • Keep client confidence out of public tools, and think carefully before entering personal or privileged material into any generative system.[3][9][10]
  • Review output as suggestions, not as an answer. A named person decides.[3]
  • Keep a record of what was proposed and who confirmed it.[3][6]

The Law Society of England and Wales has published its own guidance, Generative AI – the essentials, for solicitors considering these tools.[14] The direction of travel is already clear: the tool may assist, but the lawyer remains responsible.

About Lexiq

Lexiq is a litigation management system — case management software for litigation lawyers — built in Kuala Lumpur by Wintech Advisory Sdn Bhd. It brings court dates and limitation deadlines, files and bundles, time and billing, and client updates onto one matter. Where it uses AI to read, research or draft, the software proposes and a named lawyer reviews and confirms. It does not write to a matter on its own. Research answers are limited to authorities the firm has verified; if none are on point, it says so. Lexiq is currently in pre-pilot. Nothing in this article is a claim of certification, regulatory approval, or that using Lexiq makes a firm compliant with any circular, rule or statute.

This article is for general information. It is not legal advice, and it is not a substitute for reading the authorities cited or for professional judgment on a particular matter.

References

  1. American Bar Association, Standing Committee on Ethics and Professional Responsibility, Formal Opinion 512, “Generative Artificial Intelligence Tools” (29 July 2024). https://www.americanbar.org/content/dam/aba/administrative/professional_responsibility/ethics-opinions/aba-formal-opinion-512.pdf
  2. R (Ayinde) v London Borough of Haringey; Al-Haroun v Qatar National Bank [2025] EWHC 1383 (Admin) (6 June 2025), judgment of the Divisional Court (Dame Victoria Sharp P. and Johnson J.), at [4]–[8]. Official judgment page and PDF: https://www.judiciary.uk/judgments/ayinde-v-london-borough-of-haringey-and-al-haroun-v-qatar-national-bank/
  3. Malaysian Bar, Circular No. 242/2025 (3 July 2025), “Updates on the Use of Generative Artificial Intelligence in Legal Practice,” issued by the Bar Council Cyber and Privacy Laws Committee and the Bar Council Legal Tech, AI and Sandbox Committee, supplementing Circular No. 342/2023. https://www.malaysianbar.org.my/cms/upload_files/document/Circular%20No%20242-2025.pdf
  4. Sulok Tawie, “Chief Judge: Sabah, Sarawak High Courts to use AI to propose punishments,” Malay Mail (17 January 2020), reporting remarks of the then Chief Judge of Sabah and Sarawak, Tan Sri David Wong Dak Wah, at the opening of the Legal Year 2020. https://www.malaymail.com/news/malaysia/2020/01/17/chief-judge-sabah-sarawak-high-courts-to-use-ai-to-propose-punishments/1828985
  5. Ministry of Science, Technology and Innovation (Malaysia), The National Guidelines on AI Governance & Ethics (published September 2024). The guidelines state that they are voluntary for stakeholders and set out seven principles, including fairness; reliability, safety and control; privacy and security; inclusiveness; transparency; accountability; and the pursuit of human benefit and happiness. https://www.mosti.gov.my/wp-content/uploads/2024/09/NATIONAL-GUIDELINES-OF-AIGE-20241118.pdf
  6. OECD, Recommendation of the Council on Artificial Intelligence, OECD/LEGAL/0449 (adopted 22 May 2019; amended 3 May 2024), principles 1.1 (inclusive growth, sustainable development and well-being), 1.2 (human agency and oversight), 1.3 (transparency and explainability) and 1.5 (accountability). https://legalinstruments.oecd.org/en/instruments/oecd-legal-0449
  7. Mata v Avianca, Inc., No. 22-cv-1461 (PKC) (S.D.N.Y. 22 June 2023) (Castel, J.), Opinion and Order on Sanctions. https://law.justia.com/cases/federal/district-courts/new-york/nysdce/1:2022cv01461/575368/54/
  8. Singapore Judiciary, Guide on the Use of Generative Artificial Intelligence Tools by Court Users (applies from 1 October 2024 to matters in the Supreme Court, the State Courts and the Family Justice Courts), paragraphs 3 and 5. https://www.judiciary.gov.sg/docs/default-source/news-and-resources-docs/guide-on-the-use-of-generative-ai-tools-by-court-users.pdf
  9. Courts and Tribunals Judiciary (England and Wales), Artificial Intelligence (AI) Guidance for Judicial Office Holders (31 October 2025), replacing the April 2025 guidance. See in particular the sections on confidentiality and privacy, and on accountability and accuracy. https://www.judiciary.uk/guidance-and-resources/artificial-intelligence-ai-judicial-guidance-october-2025/
  10. Legal Profession (Practice and Etiquette) Rules 1978 (Malaysia), rule 35(b): “An advocate and solicitor shall preserve his client’s confidence and this duty outlasts his employment.” https://www.malaysianbar.org.my/article/members/laws-bc-rulings-and-practice-directions/legal-profession-practice-and-etiquette-rules-1978/rule-35-advocate-and-solicitor-not-to-abuse-confidence-reposed-in-him-by-client-
  11. Legal Profession (Practice and Etiquette) Rules 1978 (Malaysia), rule 16. https://www.malaysianbar.org.my/article/members/laws-bc-rulings-and-practice-directions/legal-profession-practice-and-etiquette-rules-1978/rule-16-advocate-and-solicitor-to-uphold-interest-of-client-justice-and-dignity-of-profession-
  12. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act), Annex III, point 8(a). https://eur-lex.europa.eu/eli/reg/2024/1689/oj/eng
  13. Rules of Court 2012 (Malaysia), Order 1A and Order 2 rule 1(2). https://www.malaysianbar.org.my/cms/upload_files/document/Rules%20of%20Court%202012.01.07.2012.pdf
  14. The Law Society of England and Wales, “Generative AI – the essentials.” https://www.lawsociety.org.uk/topics/ai-and-lawtech/generative-ai-the-essentials

A further practice direction on AI use in Malaysian court proceedings was announced by the Chief Justice in late October 2026 but had not been issued at the time of writing, so it is not cited here.