The Malaysian Bar’s International Malaysia Law Conference (“IMLC”) 2026 took place from 27 to 30 Oct 2026 at MITEC, Kuala Lumpur.
By Velary Velayathan, Low & Partners
This session asked the question now often posed, as artificial intelligence (“AI”) tools become embedded in legal practice. It was moderated by Fadhil Ihsan (Dale), a member of the Bar Council and Partner at Azwad Ihsan & Co. The panel comprised The Honourable Justice David Goddard KC, International Judge of the Supreme Court of Singapore, Singapore International Commercial Court and former Judge of the Court of Appeal of New Zealand (2019–2025); Mark Dempsey SC, an arbitrator, mediator, expert determiner and facilitator from Australia; and Sathish Ramachandran, Partner at Deol & Gill.
In response to the earlier question, Dale gave his answer at the outset: AI will not replace lawyers, but lawyers who use AI may very well replace those who do not.
The Honourable Justice David Goddard KC distinguished between replacing the tasks lawyers perform and replacing lawyers themselves. He cautioned that AI performs unevenly and has no real model of the world or common sense. Furthermore, it can yield unreliable legal research, particularly in smaller jurisdictions, as its training materials are predominantly American. As most jobs involve bundles of tasks, routine work such as document review, chronologies and proofreading is likely to be unbundled and automated, whereas judgment, wisdom, human interaction and moral responsibility will remain with humans. Drawing on the Jevons paradox and the experience with spreadsheets, he suggested that cheaper legal services may increase demand and narrow the access-to-justice gap. This will see paralegals carrying out much more legal work, as nursing practitioners now do in medicine. He urged the profession to consider the polarisation between high-end and general practice, the division of work between in-house and external lawyers, de-skilling and the breakdown of the apprenticeship model through which junior lawyers learn.
Mark Dempsey SC argued that AI may assist dispute resolvers but must not be allowed to decide. He set two conditions for its responsible use in arbitration. Firstly, the decision-making task must not be delegated. Secondly, the parties’ confidential material must be protected, which favours closed and secure legal tools over open platforms. He referred to a recent Canadian decision setting aside an award that relied on fictitious authorities, but warned that the greater future risk is an award that reads convincingly yet is not truly the arbitrator’s own, as there is no clear line between summarising evidence and evaluating it. He also highlighted data security, enforcement and personal exposure risks, and noted that assessing credibility remains a human task. Arbitrators should therefore act with the parties’ consent and take responsibility for the fact-finding, reasoning and expression of the award.
Sathish Ramachandran contended that the real danger is not that AI will think like lawyers, but that lawyers will stop thinking like lawyers and hand over their reasoning to a tool. Using AI is acceptable, he said, but intellectual outsourcing is not, and lawyers should keep their “actual intelligence” fully engaged. He placed AI within the continuum of Malaysia’s technology laws and highlighted the two circulars issued by the Malaysian Bar on the use of AI, which contain checklists of dos and don’ts, while noting that AI governance and cybercrime legislation are still developing.
On the question of what advice they would give to a lawyer starting out today, the speakers urged curiosity, a firm grounding in legal principle and reasoning, and keeping up with reliable tools. Sathish Ramachandran stressed that the first draft may come from AI, but the final draft must be the lawyer’s own. He pointed out that a disciplinary complaint will not be made against a chatbot, but the lawyer in charge. The speakers also emphasised empathy, respect for human dignity and the art of communicating with clients in ways they understand, which no machine can replace.
The key takeaways were that AI will change the tasks lawyers perform rather than replace lawyers; that judgment, accountability and human connection remain irreplaceable; and that lawyers must verify AI output and never delegate decisions. The panel’s answer to the pressing question was unanimous: No.
*This article reports on Breakout Session 5, Stream B: Legal Profession and Professional Practice.