Let’s face it. Most of our students now use AI for everything even if we all pretend otherwise—from summarizing dense readings to drafting emails, preparing for class, writing thesis proposals, entire essays, and yes, even PhD chapters. According to a 2024 study by the digital education council, 86% of students already use AI in their studies.  While AI cannot yet consistently produce work that passes on its own, a literate user—one who understands prompting and the logic of large language models—can produce results that are passable, even polished, even at the graduate level. But students are not the only ones. Many academics now rely on AI tools as well—for course preparation, for generating PowerPoints, writing emails, responding to students, and increasingly, even for marking papers. In all these uses, AI has become an infrastructural support for academic life. 

And yet, while there are clear ethical concerns tied to these practices, those concerns remain blurry, fragmented, and inconsistently articulated. For some, using AI is little more than the 21st-century equivalent of using spellcheck or Google; for others, it is a serious breach of academic integrity. Most people, though, fall somewhere in between—and the jury is still very much out to lunch. What’s striking is that for now, at least, the presence of these ethical questions does not appear to be deterring anyone. 

This raises deeper and more uncomfortable questions. If our methods of assessment and our expectations about student preparation are out of sync with what’s actually happening in real academic and professional settings—and if the skills we’re cultivating are no longer tightly aligned with what’s valued in professional practice—how should legal education evolve? This is not a problem that can be answered in the abstract. Legal education is always situated, embedded within specific contexts—jurisdictional differences, access to technology, language proficiency, AI literacy, professional cultures, and the commercial environment all shape what is possible and what is desirable. A student in a Canadian law school and one in a rural university in Egypt will experience the AI transition very differently. Still, the core questions must be asked. What should academic legal education consist of today? What counts as academic knowledge in an age where information retrieval, summarization, and even basic drafting can be outsourced to machines? How will professional certification evolve when traditional benchmarks—writing a PhD, publishing papers, mastering citations—can be reached far more quickly with the right tools? Does it still make sense to ask someone to labor over a dissertation for three years if something that looks and feels credible can be produced in twelve months or less? What is the purpose of prioritizing publications, when a seasoned academic with prompting fluency can churn out a well-structured article in a weekend? 

These are not rhetorical provocations. They are questions that demand rethinking—not just of what we teach, but of what we consider knowledge to be, and what we mean when we say someone is “qualified.”

1. What Counts as Legal Knowledge Now?

Legal knowledge has long been associated with a particular repertoire of skills: reading and interpreting texts, recalling precedent, articulating coherent arguments, and expressing these in grammatically and rhetorically competent prose. It has also been tightly linked to form—essays, case notes, doctrinal analysis, briefs, and theses—all of which have historically signified not just the student’s familiarity with legal content, but their capacity to produce this content in institutionally approved formats. However, the rise of AI in both legal education and legal practice invites us to re-examine what we mean by knowing the law. When generative tools can retrieve case law, draft arguments, summarize dense theory, and mimic citation styles with remarkable ease, the emphasis begins to shift from traditional forms of content production to something else—perhaps to evaluative judgment, critical selectivity, and most crucially, to a rearticulation of academic discernment.

Yet this is not an epistemic revolution. It is a reconfiguration—one in which long-standing forms of knowledge practice are revaluated, rather than replaced. Academic discernment has always been central to legal scholarship. The ability to tell the difference between a strong argument and a superficial one, to identify what matters in a body of case law, to recognize patterns across seemingly unrelated domains, or to spot conceptual slippages—these are not new skills. What is changing is their visibility and centrality. In an environment where machines can now do much of the assembling, retrieving, and even drafting, what remains human (at the moment) is the capacity to discern—to question the framing, to test the coherence, to notice the subtleties that fall through the algorithmic net, and, if anything, the desire to have humans perform this role.

This shift is not limited to students. Academics too are embedded in this transition. Many already use AI tools to prepare lectures, generate PowerPoints, draft emails, mark papers, and even scaffold research outputs. Academic labor is being reorganized—not replaced—through these new tools. If anything, the role of discernment is heightened: knowing what not to trust, what to discard, where to intervene, and how to shape the raw material that machines now produce at scale. In this sense, academic legal knowledge is not being hollowed out—it is being redirected toward its more interpretive, critical, and curatorial dimensions.

This evolution mirrors what is happening across the legal professions more broadly. The corporate lawyer drafting a contract can now generate a boilerplate version in minutes, tailored to jurisdiction and subject matter. But the value of the lawyer lies not in producing that initial draft—it lies in knowing where the liabilities hide, what the client’s risk appetite is, what must be bespoke. The judge who uses AI to summarize case law or identify doctrinal patterns is not thereby outsourcing her judgment; her role shifts toward distinguishing precedent that matters from one that merely appears relevant. The arbitration lawyer can feed facts and legal issues into a well-prompted AI system and receive a draft written brief in twenty minutes—but she is still the one who decides what tone to strike, which argument to lead with, and which parts to strategically omit. In each case, discernment—the ability to evaluate, to contextualize, to anticipate implications—remains central. 

And yet, this is only part of the picture. There are also strands of academic and professional life where discernment is not simply reframed, but minimized—or even rendered unnecessary. Some judges may indeed sign off on AI-generated drafts wholesale, especially in low-stakes or high-volume procedural contexts. Some law firms may come to expect that first drafts of memos or client communications be entirely AI-generated—not because the lawyers are being negligent, but because the time constraint and cost-benefit calculus make it the rational choice. In academic settings, there are already entire syllabi, slide decks, and draft papers being composed through automated tools. And again, this is not necessarily a failure of professional ethics—it may be precisely what allows an overworked academic or legal professional to meet expectations and keep the institutional machinery moving.

In such cases, discernment doesn’t vanish altogether, but it becomes selective. It is no longer a universal requirement, but a contingent one—invoked where complexity, novelty, or institutional risk demands it. For routine tasks, for repetitive forms, for standardized outputs, AI may generate results that are not only sufficient, but preferable. The idea that legal expertise consists in writing everything from scratch is increasingly untenable, not because the profession is collapsing, but because the tools are becoming normalized and institutional logics—faster turnaround, lower cost, backlog reduction—begin to favor them. 

This destabilizes not only modes of work, but the very idea of legal expertise. Historically, the stature of the legal expert—whether scholar, practitioner, or judge—rested not only on knowledge of content, but on the difficulty of acquiring and deploying that knowledge. The professional mythology of law relied on this asymmetry. Legal knowledge was complex, arcane, interpretively rich. It required years of study, training, and cultivation. And it came at a premium—epistemically, financially, socially. The figure of the jurist, whether imagined as the learned judge or the scholarly sage parsing the “right answer” from a sea of doctrine, derived authority precisely from this cultivated exclusivity.

But what happens to that mythology when someone with no legal training can obtain a reliable legal opinion from a chatbot in under five minutes? What happens when clients come to believe—not unreasonably—that paying a lawyer for an hour of research is optional, if not obsolete? When early-career academics are compared not to each other, but to generative systems that can produce publishable text on demand? These shifts don’t just affect workload or workflow. They erode the epistemic scarcity on which professional mystique has long depended. The result is a recalibration of authority. Expertise may no longer mean “I know what others do not”, but rather, “I know how to navigate, interpret, and problematize what machines can already produce”. This does not dissolve legal professionalism or academic competence, but it does displace their traditional foundations. What remains valuable is not mastery in the old sense, but a kind of epistemic choreography: knowing when to trust, when to doubt, when to slow down the automation, and when to let it run.

2. Rethinking Professional Certification: Who Gets to Say Who’s Qualified?

The transformations underway in legal knowledge and labor do not stop at how work is done—they cut directly into who is licensed to do it. Historically, professional certification in law and academia has functioned not only as a gateway to practice, but as a mechanism for sustaining the epistemic and social authority of the professions. Law schools, bar associations, and doctoral programs have long served as gatekeeping institutions—controlling not just access to legal or academic roles, but shaping what counts as legitimate knowledge, who counts as an expert, and how that expertise is measured. But if, as we saw in the previous section, large parts of legal and academic work are shifting toward routinized or automation-compatible tasks—if discernment is no longer always needed—then a fundamental question arises: why require years of study, credentialing, and training to perform work that can now be done effectively, or at least sufficiently, by machines or by individuals with minimal legal education? If the associate’s job, or the teaching assistant’s job, is increasingly to interface with systems, respond to outputs, and lightly edit automated drafts, then what does it mean to say that only someone with a JD, LLB, or PhD is qualified to perform that role?

There is a real risk that universities and professional bodies will lose their monopoly on credentialing—not because they are failing, but because the definition of “qualified” is itself being contested. Already, we can imagine private firms offering short-term, intensive induction programs tailored to very specific, low-discretion tasks in legal practice: e-discovery, contract review, litigation support, drafting procedural memos. A highly reputable firm could launch a three-month certification pathway offering hands-on training in AI tool use, document handling, and firm protocols. Graduates of that program may be just as attractive to employers—especially for entry-level or narrowly scoped roles—as someone who has spent four years studying jurisprudence, philosophy, and comparative constitutional law.

This is not dystopian; it is efficient. From the perspective of firms and institutions under pressure to reduce costs, increase turnaround, and adapt to technological realities, such private credentials may not only be acceptable—they may be preferable. Why pay for the expensive epistemic overhead of a fully trained legal scholar when the task at hand demands none of that? Why require a junior tutor in a large undergraduate law course to have a PhD, when all they are doing is grading essays largely written (in whole or part) with AI tools, or responding to emails that are themselves AI-generated? The tutor in this scenario becomes a human interface, a compliance layer, a figure of soft oversight.

But the loss here is not merely symbolic. What risks erosion in this reconfiguration is the depth and breadth that liberal arts education once provided—not as ornamental, but as formative. Legal education has never been just about professional readiness. It has been a vehicle for cultivating intellectual autonomy, critical reflection, historical memory, and ethical sensitivity. A student trained in jurisprudence or legal theory may never use those frameworks directly in day-to-day practice—but they are more likely to ask difficult questions about justice, interpretive ambiguity, institutional bias, or historical contingency. The liberal arts pre-requisite of legal education functioned as a slow-burning safeguard against technocratic narrowness. If certification becomes purely task-oriented, we risk producing professionals who can perform tasks but cannot interrogate them.

This is compounded by another danger: dependency. If entire workflows are routineized through AI systems—so much so that new professionals are trained to trust the output unless otherwise directed—then even the capacity for oversight may atrophy. The role of the human becomes supervisory, but in name only. And when supervision becomes habitual and unexamined, the very conditions for discernment begin to wither. One does not oversee what one does not understand; one does not question what one does not recognize as contingent. Moreover, the deskilling of entry-level professionals may create a structural ceiling. How will a junior associate who has never constructed a legal argument from scratch acquire the tacit judgment to challenge or improve an automated brief? How will a junior academic develop a voice in a system where originality is de-emphasized in favor of fluent synthesis? Professional development becomes hollowed out; learning becomes a kind of custodial engagement with machine outputs. The result may be a generation of legal professionals who can work faster, but not think deeper.

The mythos of expertise—the idea that one has undergone a process of intellectual and ethical formation that renders one trustworthy—does not survive this transition unscathed. Replacing it with metrics, micro-credentials, and prompt-badges creates a more agile, more modular model of competence—but also one more fragile, more susceptible to epistemic drift. The authority of the professional has always rested, at least partly, on the belief that their judgment was forged in contexts that exceed any immediate task. Remove that context, and the authority begins to resemble that of a systems operator: legitimate only insofar as the system functions. 

This is not an argument for nostalgia. It is, however, a call for clarity. If we are to move toward new models of certification—more flexible, more responsive, more plural—we must also ask what kinds of knowledge we are willing to devalue, what kinds of professionals we are prepared to produce, and what epistemic risks we are prepared to normalize. Certification is not just about qualification; it is about the stories we tell about why someone should be trusted. If professional certification is no longer the exclusive domain of universities and if legal knowledge is fragmenting into multiple layers of automation, discernment, and interface management, then what is left for universities to do? What is their role in a landscape where practical competence, task-oriented micro-credentials, and platform-based certification can bypass traditional degrees entirely? 

These are not hypothetical scenarios—they are emerging realities. And they pose a direct challenge to the historical identity of the university as both a gatekeeper of professions and a steward of deeper intellectual traditions. For much of the modern era, the university held a dual mandate: on the one hand, to prepare individuals for specific careers through structured training and assessment; on the other, to cultivate broader intellectual capacities—critical thinking, ethical reflection, historical perspective—that exceed any particular task. This was especially pronounced in legal education, which combined doctrinal knowledge with elements of philosophy, history, politics, and moral reasoning. Law was not merely taught as a set of rules, but as a living discourse shaped by cultural narratives, power structures, and ethical dilemmas. Curriculums often included courses on economics, sociology, history, and political theory, reflecting a belief that understanding law required engagement with the wider intellectual currents that inform societal organization and human behavior.

Even the most practice-oriented law degrees embedded their training within a broader academic culture, one that insisted (at least nominally) on something more than procedural fluency.

That “something more” is now at risk—not only because AI can perform many of the surface tasks once used to measure competence, but because the institutional relevance of universities is being challenged by alternative providers who promise speed, focus, and cost-efficiency. What, then, can universities offer that neither platforms nor private firms can easily replicate? The answer may lie not in abandoning the liberal and theoretical dimensions of legal education, but in doubling down on them—strategically and visibly. If discernment is becoming a selective rather than universal requirement in legal and academic labor, then universities must become the sites where discernment is cultivated as a public good. Not everyone will need deep jurisprudential reflection in their day-to-day legal work—but someone must be trained to ask whether the automation of precedent application is distorting doctrine; whether the data used in sentencing algorithms embeds bias; whether our legal vocabulary still reflects the moral and social transformations of our time.

This is not an argument for elitism, but for clarity of institutional function. If universities attempt to compete with private certification on efficiency, they will almost certainly lose. They are too slow, too regulated, too fragmented. But if they reassert their value as institutions where epistemic habits are not just trained but interrogated—where knowledge is not just used but historicized and questioned—they may retain a critical role in shaping the professions of the future, rather than merely manufacturing them.

To do this, however, universities must rethink their own pedagogical and structural assumptions. They cannot continue to teach and assess students as though AI does not exist. Nor can they continue to equate academic excellence with quantity of output, when speed and fluency are now artificially replicable. If universities are to defend and renew their role, they must take seriously the question: what cannot be trained in three months? What kind of knowledge resists automation—not because it is obscure or inefficient, but because it deals with ambiguity, judgment, pluralism, and the contestability of meaning?

There is also a political dimension. The university, at its best, functions as a counterweight to the economization of knowledge. It resists the reduction of value to utility, of insight to immediacy. If that role is to survive, it must be asserted deliberately. The very existence of low-discernment professional tracks makes the case for high-discernment spaces even more urgent. Not as luxury goods, but as forms of epistemic infrastructure without which the legal system becomes shallow, brittle, and narrowly optimized. In this light, the future of the university is neither to vanish nor to universalize itself. It is to specialize in cultivating slow, reflective, critical forms of legal knowledge that complement—but do not compete with—the automated, the promptable, and the modular. It must become a site of second-order thinking in a world increasingly saturated with first-order output. Not everyone will need that training. But someone must have it. And someone must be able to teach it.

3. The University as Place of Alternative Ordering: Reclaiming Purpose Beyond Utility

If professional certification is no longer the exclusive domain of universities, and if legal and academic work is increasingly divided between high-discretion and low-discretion tasks, then the university must ask itself not just what it teaches, but what it is. What makes it different from a training program, a certification platform, or a corporate induction course? What kind of place does it claim to be, and why should anyone enter it? 

To answer this, we must shift from thinking of the university as a neutral space—a zone where knowledge circulates—to thinking of it as a place: a structured, lived, and symbolically charged site with a particular relationship to time, authority, and subjectivity. More precisely, the university can be understood as a heterotopic place. In Michel Foucault’s sense, a heterotopia is a place that stands in contrast to the ordinary arrangements of the world. It mirrors society while simultaneously inverting or suspending its norms. It is where different temporalities operate, where identities can be reconfigured, where alternative logics are not only imagined but rehearsed. The university, at its best, has always been such a place.

In legal education, this heterotopic character has been especially pronounced. It is where students have encountered not just legal rules, but legal thought—where they could wrestle with jurisprudence, historical injustice, normative theory, or comparative frameworks in ways that the profession often lacks time or tolerance for. It is where they could argue in bad faith to test an idea, take intellectual risks without professional consequence, and encounter worldviews radically different from their own. These are not incidental features of the university—they are constitutive of its function as a place where human beings can become something other than efficient performers of tasks.

Yet this role is now endangered—not because the university has lost its value, but because the epistemic, economic, and technological forces that once deferred to it are beginning to bypass it. AI renders routine legal and academic work more efficient; credentialing becomes modular and market-driven; institutions increasingly seek measurable outputs. In such a world, the university is pressured to justify itself in transactional terms: are students employable? Are programs scalable? Is content optimized? But to accept these terms wholesale is to evacuate the very distinctiveness that defines the university as a heterotopic place—a site ordered differently, where knowledge is valued not for what it delivers, but for how it transforms.

But if the university accepts these terms too readily, it risks erasing the very distinctiveness that gives it value. It cannot outcompete the speed of platforms, the agility of private firms, or the precision of AI-driven instruction. What it can offer, and must continue to offer, is something those systems cannot replicate: a place to imagine otherwise. A place to experiment, to converse, to slow down, to inhabit uncertainty. A place to reflect not just on how to do the work, but whether the work should be done at all. A place where students encounter ideas that do not yet have application, and people they would not otherwise meet.

AI has a role in this place—but only if subordinated to those ends. Used well, AI can expand access, scaffold curiosity, and amplify learning. Used poorly, it accelerates the flattening of educational experience into a series of plausible outputs. If the university is to survive as a heterotopic place, it must insist that not all value can be measured in speed, accuracy, or optimization. It must defend the time it takes to think, the uncertainty required to learn, and the risk of failure as a necessary part of formation.

But this cannot be accomplished by nostalgia. The university cannot simply repeat the gestures of a bygone era and expect to retain its authority. It must change—not to conform to the logic of efficiency, but to better articulate its difference from it. That change begins with a sober recognition: the university’s monopoly on certification is gone, and its role as a transmitter of content is no longer exclusive. What remains is its capacity to be a place—a heterotopic place—where knowledge is not just acquired, but reimagined. Where the legal profession, and indeed society, can look to find not just trained individuals, but thinkers capable of asking what legal knowledge is for. That, in turn, demands a rethinking of the university’s structures, rituals, and pedagogies—so that its distinctiveness is not only preserved, but made newly relevant. The university must change. The next section begins to explore how.

4. The University Must Change—now.

If the university is to survive as a site where the dominant logics of utility, speed, and procedural efficiency are not simply mirrored but interrogated—then it must recognize how deeply its foundations are being shaken. The disruptions brought by AI are not merely technological; they are epistemic, institutional, and cultural. They demand not an update, but a transformation. But not all change is renewal. Some change leads to erosion, dilution, and loss. The task now is to discern what must be rethought, and what must be defended.

The most urgent shift is this: education must be reclaimed as transformative, not transactional. The university cannot be reduced to a knowledge-delivery platform, and learning cannot be reduced to credential acquisition. A legal education is not a checklist of competencies. It is a formative process that shapes how people think, argue, doubt, imagine, and judge.

This formation cannot be achieved through interface alone. It requires dialogue, friction, surprise—encounters with other minds and other ways of seeing. It requires moments of silence, boredom, failure, and risk. None of this can be outsourced to systems designed to optimize for predictability and speed.

And yet the temptation to proceduralize is growing. Increasingly, there is pressure to view education through the lens of operational efficiency: standardized content delivery, modular assessments, uniform feedback protocols. In such a view, faculty become interchangeable facilitators, and those most invested in high-discernment education—those who linger over complexity, who defend the value of ambiguity, who challenge grade inflation and resist over-scripting the curriculum—are cast as obstructive or dispensable. This is not merely a managerial error. It is an epistemic tragedy.

Tenure, in this context, is not an antiquated privilege. It is a structural commitment to the possibility of independent, unhurried, and sometimes unpopular forms of thinking. It can protect the kind of intellectual work that does not conform to performance metrics, the kind of teaching that does not scale neatly, and the kind of judgment that cannot be automated.

If the university is to retain its distinctiveness, it must protect faculty not in spite of their resistance to proceduralization, but because of it.

This also requires a shift in how we relate to students. Faculty must stop assuming that students’ use of AI tools is inherently deceptive. Often, it is adaptive. It reflects an instinct to meet demands with available tools, to synchronize with a world moving faster than we are.

The challenge is not to discipline this instinct, but to direct it—to invite students into the co-creation of new norms for learning, writing, and thinking with AI. The goal is not to punish students for adapting, but to ask with them: what is lost in automation? What remains irreducibly ours?

None of this is simple. Rethinking legal education in the age of AI means redesigning assessment, rethinking pedagogy, defending intellectual labor, and resisting the gravitational pull of procedural convenience. It means insisting that some things—conversation, reflection, judgment—cannot be scripted into rubrics or replaced by dashboards. It means slowing down in a culture obsessed with acceleration. It means allowing students to become more than users of systems—to become thinkers who can critique systems, live with uncertainty, and make judgments in contexts where rules do not suffice.

This also requires rethinking our assumptions about authorship and creativity. AI does not simply automate—it generates. It produces arguments, summaries, citations, and structure in ways that make the boundary between learning and outsourcing blurrier than ever. Unlike past tools such as citation managers or search engines, LLMs intervene in the production of meaning itself. And unlike contract cheating—where students purchase complete assignments—AI tools invite co-authorship. Students who use AI are not always avoiding work; they are often engaging with it differently, through prompting, filtering, and revision. The pedagogical question is shifting from “Did you write this?” to “How did you arrive here?” That distinction matters. It requires moving away from punitive models and toward an ethic of intellectual transparency. If creativity is no longer solely a matter of producing from scratch, but also of curating, framing, and editing, then assessment must evolve to capture those forms of engagement. The risk is not that students will cheat, but that we will fail to teach them what good use looks like.

This is not a nostalgic vision. It is a forward-looking one. Furthermore it is rooted in a commitment to the university as a place where people are formed, not processed; where education is lived, not downloaded. A place where law is not simply learned, but reimagined. A place where students are invited to reflect not only on how to work in the world, but on how the world itself might be otherwise.

And we must remember that these questions will not land evenly. The future of legal education will be shaped differently in Cairo than in Cambridge, in Accra than in Amsterdam. Universities are situated institutions. Their reinvention must be globally attentive and locally grounded.

This is not the end of academic legal education. But it is a turning point of a different order. Previous reforms have shifted curriculum, pedagogy, or access. This moment, however, threatens the epistemic foundations themselves: the authority of judgment, the value of formation, the time needed for thinking. Whether what follows is renewal or retreat will depend on how seriously we take this challenge—and how confidently we respond, not with resignation, but with re-design rooted in the clarity of purpose.