Following are the opening remarks by the Secretary for Justice, Mr Paul Lam, SC, at Hong Kong Legal Week 2026: LawTech Forum today (November 5):
 
Distinguished guests, ladies and gentlemen,
 
     Good afternoon. It gives me great pleasure to welcome you to the LawTech Forum at the Hong Kong LawTech Fest 2026.
      
     The theme of today's Forum is "Human in the Loop, Integrity at the Core - Redefining Legal Practice with LawTech". It underscores how lawyers should meet this challenge. Artificial intelligence (AI) may be powerful and, in some cases, even deliver output immaculately. It cannot, however, replace human beings, at least not for the time being, unless we allow that to happen. In legal practice, the human in the loop is not a hindrance to innovation. Rather, it is the compass that keeps innovation aligned with the pursuit of justice, the discharge of professional duties and where public interest lies.
      
     At last year's conference (LawTech Conference "The AI Era: Shaping the Legal Landscape in the 21st Century"), I advocated the "Human+AI" approach: harnessing the benefits of technology while preserving advocacy, ethical judgment, commercial acumen, critical decision-making, emotional intelligence and relationship building and management which cannot be fully automated.  
      
     One year on, AI is already transforming legal services and society. The question now is how lawyers can internalise it without compromising our professional calling.
      
     Whether you are an adviser, advocate, legislator, judge or teacher, lawyers as participants of society can help to build a sound legal framework within which technology can become all pervasive but remain safe. Once the needed protection is identified, greater certainty can be given to AI developers and users about who is responsible for what, and what remedies are available. This is where the Department of Justice (DoJ)'s work on AI governance comes in.        
      
     The Government's objective, as set out in the Policy Address 2026, is crystal clear - to achieve technology for the good of society and ensure safety and control. An effective legal framework is an essential part of that endeavour.
      
     The Department of Justice assumes a pivotal role in this process. To take forward the initiative announced by the Chief Executive in last year's Policy Address, we established an Inter-Departmental Working Group to Review Legislation to Support Wider Application of AI. Further to a Steering Committee meeting held in March this year, all bureaux and departments have completed a preliminary review of over 600 ordinances under their policy purview.   
      
     This year's Policy Address takes things forward by setting out seven aspects of AI risk governance. Amongst some of these aspects, the DoJ will play an active co-ordinating role through the platform of the Inter-Departmental Working Group, or the Law Reform Commission (Law Reform Commission of Hong Kong) may undertake independent study.  
      
     The first aspect concerns the misuse of AI to conduct criminal activities. Deepfake technology can be exploited to create pornographic images, as well as to facilitate online scams and fraud. The Sub-committee on Cybercrime of the Law Reform Commission is reviewing the existing law from the perspective of cyber-enabled crimes and will seek public views in due course. The objective is to ensure that the law is sufficiently clear and targeted to deter and punish unlawful conduct effectively.
      
     Concerns were also expressed at a recent meeting of the Legislative Council Panel on Administration of Justice and Legal Services about misuse of a person's image with deepfake which merits attention. Before going straight for the creation of any new or distinct right to protect a person's image, voice, etc, we must first carefully examine what existing legal protection there already is and consider whether a proportionate remedy could be provided under Hong Kong's legal framework when harm is done or threatened.
      
     The second aspect is the protection of social media users who are minors. A dedicated task force under the DoJ-led Inter-Departmental Working Group will co-ordinate a review of the risks and laws concerning minor's use of social media, including interaction with AI, and formulate targeted legislative and administrative measures where necessary.  
      
     While the task force will study experiences on the Chinese Mainland and overseas, it will not slavishly follow any single model. The task force will carefully consider the feasibility and effectiveness of various options in the Hong Kong context, bearing in mind any possible loopholes or unintended consequences. An effective solution may call for a cocktail of measures including not only legislation, but also codes of practice, guidelines and education to provide the flexibility necessary for a swift response, taking into account the voices of young people, parents, schools, platforms and other stakeholders.
      
     The third aspect is legal liability for accidents or damage caused by AI products. When an AI product causes harm, the apparently simple question of "Who is responsible?" may give rise to difficult issues of fault, causation, risk disclosure, traceability, explainability and human supervision. While existing laws on contract, tort of negligence, product safety, consumer protection and sectoral rules may be able to address some of these issues, questions remain as to whether they are adequate in view of the distinctive and complex features of AI products, and if not, how best the lacuna can be filled.
      
     These three priority aspects bring us back to the principle of "human in the loop". It also reminds me of the concerns recently raised by some Legislative Council members about AI hallucinations, fabricated authorities, and non-lawyers using AI to offer legal services. In response, I reiterated that the use of AI does not alter the nature of any practice to provide legal services or obviate the need to comply with the existing law. Human involvement must be meaningful and robust. This is not lip service. It requires the ability to understand risks, the readiness to intervene when necessary, and clear accountability for the decisions ultimately made.  
      
     The professional responsibility from which lawyers cannot shirk remains undiluted. An algorithm, which is built on data, looks backwards by analysing existing legislation and published judgments and tries to predict the next logical word or outcome. Nonetheless, legal practice, whether in the course of facilitating transactions or resolving disputes, calls for empathy and an understanding of the context, culture, and the subtle nuances of human behaviour.  
      
     We may also anticipate an AI-driven "armaments race" in litigation, with parties, be they legally represented or acting in person, filing ever greater volumes of material advancing apparently highly persuasive arguments supported by cited authorities, which seem too good to be on point, and evidenced by facts which seem to fit too well. While generating submissions becomes easier and quicker, one must resist the temptation to join such a race. If left unmitigated, allowing AI-generated submissions to run riot may mean more work for judges and the contesting parties to identify what is accurate and relevant to the case, hence increasing, not reducing, costs. It is important to remember how arguments are won by impactful exchanges which helpfully engage the human mind.  
      
     Turning to "integrity at the core", I cannot over-emphasise that integrity remains the foundation of clients' trust. Using AI does not alter the underlying duty of confidentiality, and other ethical rules to act in the clients' best interests and fulfil our overriding duty to the court.  
      
     To bring us to the next level, "redefining legal practice with lawtech" involves more than just equipping legal practitioners and law students with the "tech literacy" to frame appropriate questions, critically assess AI-generated materials and manage data risks. There are some long-term considerations when we seek to increase productivity and re-focus our attention.  
      
     Some of the routine work that AI can now undertake, such as research, document review and preparing drafts, used to be a fertile training ground on which generations of young lawyers developed their skill sets. If AI takes over more and more of that work and entry opportunities diminish, we risk drying up the pipeline of sufficiently experienced lawyers on whom the profession depends. The answer is not to preserve mundane work for its own sake, but to re-engineer genuine opportunities pitched at the right level for the younger generation to enter the profession, learn through supervised practice, and take on responsibilities in exercising independent judgment. If we want a capable human in the loop tomorrow, we must today continue to provide opportunities for that person to hone such skills, or even learn from mistakes.
      
     Our approach in the AI era is to embrace innovation with professional integrity and human oversight. With our common law system, bilingual legal environment and strong international connections, Hong Kong is well-placed to offer the trusted legal landscape in which both individuals and enterprises can thrive, irrespective of their legal, cultural and linguistic backgrounds.  
      
     AI may change how legal practice is conducted, but not our commitment to the rule of law. With "human in the loop and integrity at the core" close to our hearts, we have every reason to trust that lawtech serves not only the legal profession, but the wider community in a sustainable manner.
      
     On that note, I wish you would enjoy this Forum and would hear something useful to take home. Thank you.

Ends/Thursday, November 5, 2026
Issued at HKT 17:12

NNNN