The efficiency trap in law: why optimizing the past is undermining the future

Artificial intelligence is rapidly increasing the speed of legal work. Contracts are drafted in seconds, case law is summarized instantly, and due diligence that once took weeks is now completed in hours. The gains seem undeniable, and the conclusion feels almost self-evident: efficiency is improving, therefore the system is progressing.

But that conclusion rests on two assumptions that deserve closer scrutiny. First, that law is fundamentally a discipline where speed creates value. And second, that the work we are accelerating today will remain the work that matters tomorrow.

Both assumptions are increasingly fragile. Together, they form what might be called the efficiency trap: the systematic optimization of activities whose relevance is quietly declining.

Law was never about speed

To understand the trap, it helps to return to first principles. Law has never been about speed. It has always been about judgment. The ability to navigate ambiguity, interpret meaning, and anticipate consequences in situations where rules alone are insufficient. The value of legal expertise does not lie in how quickly an answer is produced, but in whether that answer is right, defensible, and appropriate to context.

The profession’s growing obsession with speed is therefore not inherent, but much more structural. It is rooted in the economics of the billable hour, where time is monetized and efficiency directly translates into margin. Within that model, faster appears better. Because it improves throughput. Artificial intelligence fits seamlessly into this logic. It accelerates drafting, research, and analysis. It reduces the time required to produce outputs. In doing so, it makes the existing system more efficient. Without questioning whether the system itself still makes sense.

Are we doing the right things?

The dominant narrative around AI in law is framed by a single question: how can we do what we already do, faster? It is the wrong question. Because it assumes that the current configuration of legal work remains valid. That the tasks, processes, and outputs that defined the profession in the past will continue to define its future.

A more fundamental question is unavoidable: Are we still doing the right things? This question becomes pressing once we acknowledge a deeper shift: the role of humans in legal work is changing.

When human roles change, work must change

For decades, the value of the lawyer was built on three forms of scarcity: access to information, the capacity to analyze it, and the technical mastery required to translate it into legal language. AI is rapidly eroding each of these. Information is abundant. Analysis is increasingly automated and drafting is becoming commoditized.

As a result, many of the activities that once defined legal expertise are losing their strategic centrality. And yet, the profession’s response has largely been to accelerate precisely those activities.

This is the second layer of the efficiency trap: we are not just moving faster. We are moving faster within a model that is becoming obsolete.

Accelerating yesterday’s logic

AI is a powerful amplifier. But it amplifies what already exists. In legal practice, this means it accelerates precedent-based reasoning, document-centric workflows, and linear processes of research, drafting, and review. These structures were not designed for today’s reality. They emerged in a context where human cognition was the primary bottleneck. Where accessing, processing, and producing information required significant time and effort. That bottleneck is now shifting.

Yet instead of redesigning work to reflect this new reality, we are optimizing the old design. The result is a form of path dependence: the future continues to be shaped by the logic of the past, even as the conditions that justified that logic disappear.

The hidden cost of efficiency

Efficiency is not the problem. What matters is what efficiency displaces. As speed becomes dominant, three essential elements of legal practice begin to erode.

Friction is the first. What appears inefficient is often where thinking happens, where assumptions are challenged and alternatives are explored.

Deliberation is the second. Legal reasoning requires time, not as a constraint, but as a resource. Compressing time compresses thought.

Reframing is the third, and perhaps the most important. Speed reinforces a problem-solving mindset: take a defined issue and produce a solution quickly. But many of today’s challenges are not well-defined problems. They require the problem itself to be questioned.

When everything becomes faster, that space disappears.

The illusion of progress

The result is a subtle but powerful illusion. When a contract is produced in seconds instead of hours, it feels like progress. When research becomes instantaneous, it feels like advancement. But these gains are largely operational, not strategic.

They improve how work is done without asking whether the work itself remains relevant.

This pattern is not unique to law. In periods of technological change, organizations often invest heavily in optimizing existing processes, while underinvesting in rethinking them. In stable environments, that is rational. In periods of structural transformation, exactly this kind of reasoning becomes a liability.

Innovation research has long shown that meaningful progress does not come from refining existing systems, but from challenging their underlying assumptions. There is a persistent contradiction here. We expect new outcomes, while relying on old structures. We want forward-looking solutions, while accelerating backward-looking processes. No amount of efficiency can resolve that tension.

Redesign, don’t optimize

As AI takes over more analytical and drafting work, the comparative advantage of human lawyers must shift. Not toward doing the same work faster, but toward doing fundamentally different work.

This means moving from problem solving to problem framing. It means defining what the real issue is, rather than accepting it as given. It means integrating legal, business, ethical, and societal perspectives. It means understanding second- and third-order effects, and developing approaches that do not yet exist in precedent.

This is not an incremental extension of legal work. It is a different mode of thinking.

The implication is straightforward, but not easy. Law firms must move from optimizing workflows to redesigning work. Efficiency gains should not simply reduce cost. They should create space for deeper and different thinking. Teams should not just execute faster. They should be structured to bring in broader perspectives. Value should not be measured in speed or hours, but in the quality of insight and judgment. This is a shift not just in tools, but in mindset.

The hardest part is not adopting new technology. It is letting go of the assumptions embedded in the old model. Those assumptions are deeply rooted, think of billing structures, in career paths, and in professional identity. Efficiency improvements are attractive because they leave these foundations intact. But the transformation required now is not additive. It is substitutive.

It requires accepting that improving the current way of working is no longer enough.

Strategy instead of speed

The legal profession is not at risk of moving too slowly. It is at risk of moving quickly in the wrong direction. AI will continue to improve efficiency. That is inevitable. The real question is what we do with that efficiency.

Do we use it to reinforce the past? Or to rethink the future?

Because when the role of humans changes, the work itself must change. And in that moment, the central challenge is no longer speed. It is direction and in other words: strategy.