A senior associate at an international firm gains access to a new AI-powered contract review tool. She attends the training, watches the demo, and nods along. Then she goes back to her desk and reviews the next contract exactly as she has for the past twelve years: clause by clause, manually, from start to finish. The AI tool sits open in another tab, unused.
What is the main issue? Simply this: everything she knows about doing her job well is telling her to keep doing it the old way. And that instinct, which served her perfectly for years, is now the main obstacle standing between her and a better workflow.
This is the unlearning problem. And it is, quietly, the biggest barrier to AI adoption in legal practice today. Not the technology. Not the budget. Not regulation. The barrier is what professionals already know, and their reluctance to let parts of it go..
What Unlearning Actually Requires
Unlearning differs from forgetting. Rather than erasing years of experience or discarding legal expertise, it means recognising that certain habits, reflexes, and assumptions that once made sense no longer serve you in a changed environment, and then choosing to act differently.
In legal work, the list of things worth unlearning is specific. The belief that hours spent equals value delivered. The instinct to control every step of a process manually, because that is the only way to guarantee quality. The assumption that delegating cognitive tasks to a machine is inherently risky, while delegating them to a tired junior at 2 a.m. is just how the profession works. The reflexive “we have always done it this way” that shuts down conversations before they start.
These are deeply embedded professional patterns, built up over years of training and reinforced by institutional culture. That is precisely what makes them hard to change. Unlearning asks you to question things that feel like part of your professional identity, and that is uncomfortable. Acknowledging that friction is the first step. Pretending it does not exist is how adoption stalls.
The New Mindset (and Why It Is Worth the Discomfort)
What replaces the old habits is a different kind of professional skill: the ability to think strategically about where your judgment matters most, and to let go of the tasks where it does not.
The new mindset is about asking better questions. Not “how do I review this contract?” but “what is the most efficient way to surface the issues that matter in this contract?” Not “did the AI get this right?” but “what would the AI likely miss, and where should I focus my review?” This kind of thinking is harder than manual review. It requires a deeper understanding of both the legal matter and the tool being used.
The benefits are real. work becomes more interesting. And firms that support this transition see measurable results: faster turnaround, more consistent output, and teams that can handle higher volumes without burning out.
But let us be honest about the difficulties too. The fear of becoming irrelevant is real, especially for professionals whose value proposition has always been doing the work rather than directing the work. There is also a generational dimension: younger lawyers often adapt faster, which can create tension in hierarchical teams. Lastly, there is a genuine risk of over-correction, where teams move too quickly toward AI dependence and lose the deep expertise that makes human oversight meaningful.
The Balance Sheet of Unlearning
Any honest conversation about unlearning has to weigh both sides. On the positive side: professionals who actively unlearn outdated methods gain a competitive edge. They adapt faster, deliver more value per hour, and position themselves as the kind of lawyers clients increasingly want, the ones who use every available tool to get to the right answer (and not just the familiar one). Their firms become more attractive to talent that does not want to spend the first five years of a career on work a machine can do in minutes.
On the other side: unlearning done poorly creates its own problems. Abandoning manual skills too quickly means losing the ability to catch errors when AI fails (and it will fail). Pushing change without adequate training breeds resentment. And there is a real risk that institutional knowledge, the kind that lives in experienced heads rather than in databases, gets lost in the transition if it is not deliberately preserved.
The right approach is deliberate and staged. Identify the specific habits and processes that are holding your team back. Provide the training and support that makes change feel achievable rather than threatening. Measure what improves and what does not. And keep enough of the old expertise alive that you can fall back on it when the technology stumbles.
Moving Forward by Letting Go
The legal profession has always been conservative about change, and for good reason. Precedent matters. Process matters. Rigour matters. But there is a difference between being rigorous and being rigid. The firms and legal departments that will lead in the next few years are the ones that can tell the difference, that can hold on to the professional standards that define good legal work while releasing the habits that merely define how things used to be done.
Learning AI is the easy part. Courses, demos, pilots: those are everywhere. The hard part is unlearning the assumptions that make those tools feel unnecessary.
Start there.
At Better Ipsum, we help law firms and corporate legal departments navigate the human side of AI adoption, from mindset shifts to operational redesign. If your team is ready to move forward but struggling to let go of what is holding it back, let’s talk.