How artificial intelligence and democratised access to legal information are dismantling the knowledge monopoly that has long sustained the prestige, billing structures, and social authority of Singapore’s legal profession
Abstract
For two centuries, the legal profession in Singapore has derived its social authority and economic power from a structural monopoly over legal knowledge — a monopoly enforced not merely by statute but by the sheer difficulty of accessing, interpreting, and navigating the law. That monopoly is now under unprecedented pressure. The convergence of generative artificial intelligence, the expansion of open-access legal information infrastructure, and deliberate state-led democratisation efforts is collapsing the informational asymmetry on which legal privilege has historically rested. This article examines the historical foundations of that privilege in Singapore’s common law tradition, traces the institutional and technological forces now eroding it, and analyses the uneven consequences for different tiers of the profession. It argues that the erosion of legal privilege through open knowledge is neither complete nor uniformly emancipatory: it threatens to hollow out the professional middle while intensifying competition for the irreducibly human skills that remain resistant to automation. The implications extend beyond the profession to the architecture of access to justice in Singapore.
I. The Architecture of Legal Privilege in Singapore
1.1 Knowledge as the Source of Professional Authority
The sociology of professions has long recognised that the peculiar social authority of lawyers rests not simply on competence but on the controlled scarcity of knowledge. Andrew Abbott’s foundational account in The System of Professions (1988) argued that professions secure their position through ‘jurisdictional claims’ — the assertion of exclusive mastery over abstract knowledge that others cannot easily replicate. Law is among the oldest and most successful examples of this model. Its language is deliberately opaque. Its procedural machinery is labyrinthine by design. Its knowledge base — statutes, case law, subsidiary legislation, regulatory codes — was historically impenetrable without years of institutional formation.
Singapore’s legal system, rooted in the Second Charter of Justice of 1826 and two centuries of common law development, has inherited and refined this architecture. Legal education reflected these assumptions with particular clarity. As Ben Chester Cheong, writing in The Straits Times in February 2026, recalled of his own legal training: examinations were closed book, students spent long hours memorising cases and statutes, and access to knowledge itself formed part of professional value. In practice, this translated to labour-intensive processes — weeks reviewing documents in physical data rooms, billing structures that monetised the very difficulty of information retrieval.
The economics followed the epistemology. When knowledge is scarce and access to it expensive, whoever controls that access commands premium rents. Singapore’s legal market has reflected this logic: large firms charging substantial hourly rates for work that was, in significant part, the conversion of inaccessible information into usable professional advice. The lawyer was, functionally, a human API — an interface between a client’s problem and a body of law the client could not independently navigate.
1.2 Institutional Gatekeeping: LawNet, Subscription Access, and the Price of Knowledge
The institutional architecture of legal knowledge in Singapore has historically reinforced this scarcity. LawNet, administered by the Singapore Academy of Law, has served since 1988 as the primary repository for Singapore case law, statutes, subsidiary legislation, and legal commentary. It is, in the characterisation offered by one commentator, ‘essentially a subscription site.’ Its pay-per-use rate has been SGD 57.53 for the first thirty minutes — the equivalent, as one blogger noted pointedly, of nine Filet-o-Fish meals — rising to SGD 8 for every subsequent half hour. Standard monthly subscriptions have run to approximately SGD 206 per month.
Free and open access to law matters because litigants-in-person are already at a disadvantage if their counterparty has counsel — we do not need to exacerbate this inequality by making them worse-off in terms of access to legal information too.
This was not a failure of design but a reflection of it. The financial model of legal knowledge infrastructure encoded the profession’s informational monopoly in the price of access itself. Self-represented litigants — a population whose significance has been documented in the first academic study of their kind in Singapore, by Associate Professors Jaclyn Neo and Helena Whalen-Bridge — faced a double disadvantage: the absence of professional representation and the cost of accessing the case law on which they might base their own arguments. The courts presupposed lawyers; the information infrastructure presupposed subscribers.
Access to court judgments remained partially open: the Supreme Court and State Courts published judgments online for the preceding three months at no cost, and Singapore Law Watch aggregated Court of Appeal and High Court decisions on a similar rolling basis. SAL also maintained an open-access repository of Supreme Court judgments from 1965. But comprehensive, searchable, archived access — the kind that professional legal research requires — remained behind a paywall, accessible in full to NUS Law students and staff, and to paying subscribers in the profession.
II. The Forces of Democratisation
2.1 Generative AI and the Collapse of Informational Asymmetry
The entry of large language models and generative AI into legal research has fundamentally altered the economics of legal knowledge retrieval. Tools like Harvey AI — now deployed by Singapore’s Small Claims Tribunals and adopted by major Singapore firms including Rajah & Tann, WongPartnership, and several other Big Four practices — can synthesise case law, draft documents, review contracts, translate court correspondence into multiple languages, and provide preliminary assessments of case outcomes in fractions of the time previously required of junior associates.
The productivity implications are arresting. Generative tools have cut document summary time ‘from approximately two days to roughly ten minutes’ in documented Singapore deployments. For legal research tasks that once required a qualified lawyer billing at several hundred dollars an hour for several days, AI can now produce functional equivalents in minutes at marginal cost. The informational asymmetry that sustained the premium on legal knowledge retrieval is not simply narrowing — it is collapsing.
At the TechLaw.Fest 2025 conference, the Singapore Academy of Law unveiled LawNet 4.0 — a generative AI-enhanced version of Singapore’s primary legal research platform. By combining large language model capabilities with curated, authoritative legal data, LawNet 4.0 represents a deliberate state-led effort to integrate AI into the core infrastructure of legal knowledge access. Singapore High Court Judge Justice Kwek Mean Luck’s announcement of the platform was described by observers as a signal that ‘AI is moving from experimental pilots to core national infrastructure.’ The knowledge fortress is being rewired from within.
2.2 Self-Represented Litigants and the Emerging AI-Assisted Public
Generative AI has also been deployed directly at the interface between the legal system and the lay public. In Singapore’s Small Claims Tribunals, Harvey AI has been used to help self-represented litigants answer legal queries, prepare their cases for hearing, and receive preliminary assessments of likely outcomes. The Singapore courts have likewise issued guidance on the use of generative AI by court users — applicable to both lawyers and self-represented persons — signalling that AI-assisted legal engagement is now an anticipated and normalised feature of the justice system rather than an exceptional deviation from it.
This represents a categorical shift in the structural relationship between the profession and the public. Self-represented litigants in Singapore have historically faced formidable disadvantages: the asymmetry of legal knowledge documented by Neo and Whalen-Bridge, the procedural complexity of an adversarial system designed around the assumption of legal representation, and the financial barriers of information access. AI tools do not eliminate these disadvantages, but they materially alter the informational component. A litigant who can access a generative model capable of synthesising relevant case law, explaining procedural requirements, and drafting coherent pleadings has meaningfully more parity with a represented counterpart than was possible a decade ago.
The Singapore Management University’s research on AI in Singapore’s judiciary, published in 2025, documents this transformation in granular institutional detail — recording AI use cases across case summarisation, evidence review, and litigant assistance, including tools like the Divorce Assets Informative Division Estimator. The judiciary’s posture, as articulated by Chief Justice Sundaresh Menon, is explicitly one of using AI to expand access to justice, not to preserve professional barriers. The public interest framing is doing quiet but significant work on the profession’s monopoly.
2.3 State Strategy: National AI Strategy 2.0 and Budget 2026
Singapore’s democratisation of legal knowledge is not accidental. It is the product of deliberate state strategy. The National AI Strategy, first published in 2019 and updated in 2023 as NAIS 2.0, positioned Singapore as a jurisdiction committed to AI adoption across professional sectors with explicit access-to-justice dimensions. The Ministry of Law’s Legal Technology Platform, developed in collaboration with legal technology firm Lupl, provided law firms with subsidised access to AI-enabled case management and research infrastructure.
The signal from Budget 2026, delivered by Prime Minister Lawrence Wong on 12 February 2026, was unambiguous. Explicitly targeting legal and accountancy professionals as the first wave of a broader AI reskilling initiative, Wong stated: ‘Because AI is already reshaping many forms of white-collar and cognitive work, we will start with the accountancy and legal professions, and progressively extend them to other fields.’ The budget included AI literacy courses for legal practitioners, six months of free access to premium AI tools for those completing training, and 400 percent tax deductions on qualifying AI expenses for enterprises. The state is not merely permitting the erosion of legal privilege through open knowledge — it is actively accelerating it.
Our advantage does not lie in building the largest frontier models. It lies in deploying AI effectively, responsibly, and at speed. — Prime Minister Lawrence Wong, Budget 2026
This state strategy reflects a broader political economy. Singapore, as Wong noted, faces ‘structural constraints — limited natural resources, a rapidly ageing population, and a tight labour market.’ In this context, AI’s capacity to expand the productive frontier of professional services is of direct national interest. That this simultaneously erodes the informational monopoly of legal professionals is, from the state’s perspective, a feature rather than a bug.
III. The Consequences for Legal Privilege
3.1 The Commoditisation of Junior Legal Work
The most immediate consequence of open knowledge democratisation in Singapore is the commoditisation of the work historically performed by junior lawyers. Research memos, document review, contract drafting, correspondence summarisation — the tasks that constituted both the economic foundation of large-firm associate billing and the training ground for the next generation of senior practitioners — are precisely the tasks most amenable to AI automation. The Law Society of Singapore president, Professor Tan Cheng Han, explicitly acknowledged at the Opening of the Legal Year 2026 that ‘AI will likely reduce demand for routine legal work typically undertaken by junior lawyers.’
The human cost of this process is already visible in Singapore’s attrition data, which predates AI but has been sharply accelerated by it. Surveys conducted at the 2024 and 2025 mass call ceremonies — the formal entry point to Singapore’s legal profession — found that approximately 60 percent of newly qualified lawyers planned to leave legal practice within the next five years. Chief Justice Menon, responding to these findings at the Opening of the Legal Year on 12 January 2026, expressed concern that ‘legal practice may no longer be seen as a viable long-term career for many.’ The SAL’s own data showed accelerating transitions from private practice to in-house roles, concentrated among lawyers with two to four years of post-qualification experience — precisely those whose work is most automatable.
The attrition pre-dated AI and was driven by a constellation of factors: unreasonable work hours, toxic workplace culture, inadequate compensation relative to alternative careers, and poor work-life balance. But AI introduces a qualitatively new dimension to the problem. The junior lawyer’s position in the traditional pyramid was always partly justified by economic necessity — firms needed bodies to perform high-volume, low-complexity work, and the junior associate was that body. As AI performs this work more cheaply and rapidly, the economic rationale for the junior-heavy model evaporates, and with it the training pipeline that has historically produced Singapore’s senior legal talent.
3.2 The Differential Erosion of Privilege Across Firm Tiers
Legal privilege in Singapore is not a uniform phenomenon. It is stratified across firm tiers — the large full-service firms at the apex of the market, a middle tier of generalist and specialist practices, and solo practitioners and small firms serving individual and small business clients. The erosion of privilege through open knowledge is affecting these tiers very differently.
For large firms, AI represents both a threat and an opportunity. BDO’s analysis of Singapore’s legal market in October 2025 found that ‘legal firms face intense competition and margin pressure, with clients more value-conscious and moving toward fixed-fee models.’ AI compresses billable time in research and drafting, threatening revenue in the short term. But large firms also have the capital to invest in proprietary AI infrastructure — Allen & Gledhill, for instance, partnered with Singapore GenAI startup Pand.ai to deploy a privately managed, on-premises large language model — creating competitive advantages through data quality and tool sophistication that smaller competitors cannot easily replicate.
For the middle tier, the picture is considerably bleaker. BDO’s survey found that 66 percent of smaller Singapore law firms cited budget constraints as a barrier to technology adoption. The productivity gains of AI flow disproportionately to those who can afford the most sophisticated tools — a dynamic that mirrors the Harvard Law Review’s warning about AI amplifying rather than narrowing inequality in legal services. The informational privilege that once gave the competent generalist practitioner a defensible market position is dissolving; the capital advantage that might replace it is out of reach.
For solo practitioners, the disruption is most ambiguous. On one hand, AI tools give a single-lawyer practice access to research and drafting capabilities previously available only to firms with deep associate benches. On the other hand, the same tools are simultaneously lowering the threshold for self-representation, shrinking the client population that needs professional assistance for simpler matters.
3.3 The Credential Under Pressure
The erosion of legal privilege through open knowledge puts structural pressure on the credentialling function of the bar examination and the law degree. If the social function of these credentials was partly to certify mastery of an information base that was genuinely hard to access — and the economic returns to the credential partly reflected the scarcity of that mastery — then AI weakens the rationale for those returns. The Ministry of Law’s September 2025 public consultation on its draft Guide for Using Generative AI in the Legal Sector acknowledged the fundamental shift: it mandated a ‘lawyer in the loop’ principle and required that AI-generated outputs be verified by legal professionals, preserving the credentialling function in a supervisory rather than productive role. But this is a regulatory holding position, not a permanent settlement.
Chief Justice Menon, addressing this at the Opening of Legal Year 2026, identified legal education reform as one of the four priority workstreams for the new committee co-led with Law Minister Edwin Tong. Law schools, he stated, must revise their curricula to reflect modern practice and technological realities. The implicit acknowledgement is that what legal education has historically been producing — lawyers credentialled for an information-intensive model of practice — is not what the profession will require in the decade ahead.
3.4 The Paradox: Democratisation Without Equity
There is a temptation to read the erosion of legal privilege through open knowledge as a straightforwardly progressive development — the dismantling of an exclusionary professional fortress and the expansion of meaningful access to justice. The reality is more complicated, and the Singapore context illustrates the paradox with particular clarity.
The open knowledge developments described here do materially improve the position of self-represented litigants and clients who previously could not afford professional legal assistance. The deployment of Harvey AI in Singapore’s Small Claims Tribunals, the AI-assisted tools for self-represented parties in the Supreme Court, and the broader availability of high-quality generative models for legal research represent genuine democratising forces. These are not trivial gains: they affect the everyday legal interactions of ordinary Singaporeans navigating tenancy disputes, employment claims, family proceedings, and consumer matters.
But the same forces that are democratising access to basic legal information are also concentrating advanced legal capability among those who can afford the most sophisticated AI tools. The client who retains a large Singapore firm using proprietary, domain-trained AI models against a self-represented litigant using a general-purpose AI chatbot is not in conditions of parity. The knowledge gap has narrowed; the capability gap may have widened. Access to justice is not simply a function of access to information — it is also a function of the quality and sophistication of the tools deployed to convert that information into effective advocacy.
This is the deep structural tension in Singapore’s current moment: a state that is simultaneously committed to democratising legal access and to maintaining Singapore’s position as Asia’s premier legal hub. These are not incompatible goals, but they are in tension when the capital requirements for elite AI infrastructure exceed the reach of the public interest sector. The Ministry of Law’s Productivity Solutions Grant for the Legal Sector, and the Legal Innovation and Future-Readiness Transformation pilot, offer partial responses — subsidising technology adoption for smaller firms. Whether these instruments are sufficient to prevent the polarisation of legal capability along the lines of AI investment remains an open question.
IV. What Remains: The Irreducible Bases of Legal Authority
4.1 The Judgment Premium
Not all of legal privilege is reducible to informational monopoly. The erosion documented in this article is real and consequential, but it is also partial. The work that AI performs well — research synthesis, document review, drafting from templates, translation, pattern recognition across large case corpora — is the informational substrate of legal practice. The work that AI performs poorly or not at all — the exercise of professional judgment under conditions of genuine uncertainty, the navigation of relational complexity in high-stakes negotiations, the advocacy that persuades rather than merely informs, the ethical reasoning that protects a client’s long-term interests against their short-term preferences — these capacities do not depend on informational scarcity.
Chief Justice Menon has consistently articulated this distinction in his public addresses. AI tools have ‘proved their ability to outperform lawyers at times in areas such as summarisation of documents and analysis of court transcripts,’ but they cannot ‘replace the professional judgment and ethical duty of lawyers.’ The Senior Parliamentary Secretary for Law, Eric Chua, restated the position at TechLaw.Fest 2025: ‘Generative AI may be able to draft documents or analyse cases, but it cannot replace the professional judgment and ethical duty of lawyers.’ The regulatory and institutional architecture of Singapore’s legal profession — the bar examination, the professional oath, the disciplinary jurisdiction of the Law Society — preserves the accountability structures that give professional judgment its social warrant.
What this means in practice is that the premium on legal work is being redistributed rather than abolished. Work that was previously valued because it required access to scarce information is devaluing. Work that is valued because it requires experience-shaped judgment, institutional trust, and human accountability is, relatively speaking, appreciating. The profession is being restructured around a narrower but arguably more authentic basis for its authority.
4.2 Trust, Accountability, and the Relational Dimension
Legal privilege has never rested on knowledge alone. It has also rested on trust — the client’s trust that the lawyer’s advice is given in their interest, that confidentiality will be maintained, that the lawyer will be accountable for errors in ways that an AI system cannot be. These relational and accountability dimensions of legal authority are not reducible to information, and they are not easily replicated by AI tools.
The Singapore courts’ guidance on generative AI use — requiring verification of AI outputs, imposing responsibility on the human user, declining to make AI use a pre-emptive declaration requirement but insisting on truthful disclosure when asked — reflects a considered attempt to preserve the accountability structures of professional practice while allowing productive AI adoption. The ‘lawyer in the loop’ principle is, at one level, a regulatory artefact protecting professional jurisdiction. At another level, it reflects a genuine recognition that legal accountability structures serve public functions that market forces alone will not preserve.
V. Conclusion: A Profession Restructured
Two centuries after the Second Charter of Justice established Singapore’s legal system, the profession is navigating a transformation that goes to the roots of what legal privilege is and why it has commanded the deference it has. The erosion documented in this article is not the destruction of the legal profession — it is its restructuring on a different informational foundation.
The knowledge monopoly that sustained closed-book examinations, premium billing rates for research and document review, and the gatekeeping function over access to legal information is dissolving under the combined pressure of generative AI, open-access legal infrastructure, and deliberate state strategy. Singapore’s own institutions — LawNet 4.0, Harvey AI in the tribunals, Budget 2026’s AI training mandates, the Ministry of Law’s GenAI guidance — are agents of this dissolution as much as its subjects.
The consequences are unevenly distributed. Junior lawyers face the most immediate threat, their economic function in the traditional pyramid eroded by automation. The professional middle risks hollowing out as capital requirements for AI infrastructure exceed the reach of smaller practices. Large firms face margin compression but possess the resources to adapt. Self-represented litigants gain meaningful access to legal information but not necessarily the capability advantage that sophisticated AI tools provide to well-resourced parties.
What survives this restructuring — what cannot be automated, democratised, or commoditised — is the irreducibly human dimension of legal practice: the judgment that comes from experience, the trust that comes from accountability, the advocacy that comes from understanding what is at stake for the person across the table. These capacities have always been the deepest basis of legal authority. The erosion of privilege through open knowledge does not diminish them. It strips away what was never genuinely theirs to claim, and leaves what was.
The question Singapore now faces — and it is as much a political as a professional question — is whether the benefits of this restructuring will be distributed equitably enough to constitute genuine progress, or whether the democratisation of legal information will coexist with a new and more technologically-mediated form of legal inequality. The answer will depend on investment choices, regulatory decisions, and the willingness of the profession’s own institutions to reimagine their purpose in a landscape where the fortress of knowledge has open gates.
References and Sources
Abbott, A. (1988). The System of Professions: An Essay on the Division of Expert Labor. University of Chicago Press.
BDO Singapore. (October 2025). Reinventing Law: Singapore’s Legal Professionals in Transformation.
Chambers and Partners. (2025). Artificial Intelligence 2025 — Singapore. Global Practice Guides.
Cheong, B.C. (February 2026). ‘As AI takes on legal work, what happens to the prestige of being a lawyer?’ The Straits Times.
GovInsider. (2025). Four ways AI is shaking up Singapore’s legal practice.
Heidoh. (January 2026). Singapore’s legal profession to tackle AI disruption and attrition as judiciary marks bicentennial.
Ministry of Law Singapore. (September 2025). Public Consultation on Guide for Using Generative Artificial Intelligence in the Legal Sector.
Neo, J. & Whalen-Bridge, H. (2021). Litigants in Person: Principles and Practice in Civil and Family Matters in Singapore. Academy Publishing.
Nucamp. (September 2025). Will AI Replace Legal Jobs in Singapore?
Oxford Institute of Technology and Justice. (2025). Singapore — Issuing a National AI Strategy to Harness AI’s Potential.
Remolina, N. (July 2025). AI in the Judiciary: The Singapore Case. Singapore Management University School of Law Research Paper. SSRN 5367843.
Singapore Academy of Law. (October 2025). TechLaw.Fest 2025: A Festival of Firsts.
Singapore Budget 2026. Prime Minister Lawrence Wong. Budget Statement, 12 February 2026, Section C: Harness AI As A Strategic Advantage.
Singapore Law Watch. (January 2026). Law Minister and Chief Justice to Co-Lead Committee to Tackle Lawyer Attrition Rates, Impact of AI.
Tomoe.asia. (January 2023). Getting Judgments for Free in Singapore, or: It Shouldn’t Cost 9 Filet-o-Fish Meals to Access State Court Judgments on LawNet.