The same promise echoes at every legal technology conference. AI automates the tedious tasks and allows lawyers to focus on what really matters. While this is a seductive vision, it contradicts the best research we have about what actually happens when knowledge workers deploy these tools.
Important points:
-
-
-
The promises about productivity are deeply false. — The latest research shows that AI enhances jobs, not reduces them. Lawyers are increasing their work speed, taking on broader responsibilities, and extending their hours without realizing the expansion. Additionally, AI prompts feel more like small talk than labor, allowing lawyers to slip work into evenings and weekends without registering as additional effort.
-
Self-reinforcing acceleration is the real risk — AI speeds up tasks, increases expectations, increases dependence, expands scope, and ultimately creates a cycle that causes burnout in professions already plagued by AI.
-
Purposeful integration is the antidote — Legal organizations must promote intentional governance structures that consider how people actually behave toward AI, rather than leaders imagining how people will or should behave.
-
-
welcome home AI law professor. Last month, I looked at how AI is forcing us to. Rethinking the training of young lawyers. This month, we examine questions that affect all lawyers. What happens when the promised efficiency gains don’t materialize as expected? A recent study from the University of California, Berkeley suggests that the answer is trickier than most law firm leaders realize.
If you’ve attended a legal technology conference in the past two years, you’ve probably heard the following sales pitch: Automate the mundane and increase the meaningfulness.
a research has been published [subscription needed] in harvard business review Aruna Ranganathan and Shin-Chi Maggie Yeh, researchers at the University of California, Berkeley, suggest that we should be more skeptical. They tracked how generative AI (GenAI) changed work habits over eight months at a 200-employee technology company. Their findings were surprising. AI tools did not reduce jobs. Rather, they strengthened that.
The study found that tech workers surveyed work faster, take on a wider range of responsibilities, extend their work hours into evenings and weekends, and multitask more aggressively, even when they’re not asked to do so. The promise of liberation became a reality of acceleration and overwork.
For those of us in the legal profession, this should be a wake-up call.
3 forms of reinforcement
Researchers identified three patterns that will be familiar to anyone who watches lawyers implement GenAI into their business processes.
Expanding tasks
AI is filling knowledge gaps, allowing experts to take on responsibilities that previously belonged to others. Product managers started writing code, and researchers took on engineering tasks. In the legal context, the similarities are also clear. Employees use AI to challenge tasks that were reserved for senior lawyers. Paralegals draft documents that previously required the supervision of an attorney. Sole practitioners tackle problems outside of their core expertise because AI tools make them feel more manageable. As a result, rather than less work being distributed more efficiently, more work is concentrated among fewer people, and there is less organizational knowledge to guide outcomes.
blurred border
AI has blurred the line between work and non-work. Because facilitating AI is more like chat than work, lawyers (like the tech workers in the study) may slip work into their lunch breaks, evenings, or commutes without registering it as additional effort. Conversational interfaces are appealing because they don’t feel like work. But it’s a job and more than that.
Popular multitasking
Workers managed multiple AI threads simultaneously, creating a sense of momentum that masked the increased cognitive load. For lawyers, this means running research queries in parallel, producing multiple documents at once, and constantly monitoring the output of AI, all while believing they are saving time.
productivity trap
The most important insight from the research is that these effects are self-reinforcing. AI accelerates tasks and increases expectations for speed. As speed increases, the reliance on AI increases, and with increased reliance on AI, the range of things humans can experiment with increases. And as the scope expands, there’s even more work involved. Rinse and repeat.
Parkinson’s Law: “Work will expand to fill the available time for completion.”
In a profession already plagued by burnout, this cycle should be a wake-up call for us. The adoption of AI in the legal industry is largely driven by the promise of being able to do the same work in less time. But if the Berkeley study is any guide, what’s actually happening is that we’re doing more research. Same amount More time or more work more While telling myself that I’m becoming more productive.
And because the extra effort feels voluntary, committed leadership may not notice the problem until it manifests as a mistake, exhaustion, or ethical flaw. In law, the cost of poor judgment is not just missed deadlines, but the client’s freedom, livelihood, or life savings.
From productivity to purposeful practice
Berkeley researchers are proposing what they say is. AI practice It consists of intentional norms and routines that structure how AI is used. This includes deciding when to stop and how the work should or should not be scaled up. I would like to go further. For legal organizations, purposeful AI integration requires more than workplace health standards. It requires a strategic framework that aligns AI capabilities with an organization’s mission, ethical obligations, and sustainable human performance.
This means, first of all, being honest about what the AI is actually doing to your workloads, rather than what you expect it to do. If your company implemented AI with the expectation that it would reduce employee hours, audit whether that actually happened or if employees were simply filling the recovered time with more work.
Second, it means building governance structures that take into account how people behave. actually take action Use these tools instead of imagining how leaders would do it. Berkeley’s investigation found that workers voluntarily expanded their workloads without guidance from management. Top-down AI policies that focus only on permissible uses will miss the visible and possible escalation.
The most important insight from the research is that these effects are self-reinforcing. AI accelerates tasks and increases expectations for speed. As speed increases, the reliance on AI increases, and with increased reliance on AI, the range of things humans can experiment with increases. And as the scope expands, there’s even more work involved.
Third, it means making space for distinctly human tasks that AI cannot replicate, such as judgment, empathy, ethical reasoning, and creative problem-solving that comes from real human interactions rather than conversations with chatbots. Researchers also found that AI-powered work is becoming increasingly lonely, continuous, and dangerous.
The story that AI will enable lawyers to take on higher-value jobs is more than just optimism. It’s a misunderstanding of how these tools interact with human psychology. AI does not create leisure. capacity is created — and without intentional structure, that capacity ends up being filled with more things than strategic thinking. all.
It is clear that AI will transform the legal profession, but the real question is whether law firms will integrate it. with a purposeshaping it to serve the well-being of their values, clients, and professionals. Or whether they will quietly allow technology to shape us into something we never intended.
