AI can accelerate the work. Who will develop the judgement?
In professional services, the path from junior colleague to trusted adviser has rarely been set out in a manual. People learned by doing: researching a market, checking an assumption, preparing a first draft, listening to a client conversation, observing senior colleagues. Much of this work was routine. It was also where professional judgement began to form.
AI is changing that path. It can produce a credible market overview in minutes, suggest interview questions, summarize a lengthy document or prepare an initial analysis. Used well, it gives teams more time to focus on clients and complex decisions. Yet the work it makes faster often overlaps with the work through which younger professionals used to learn.
The question for leaders is whether their organizations are still creating the experiences that turn promising people into experts – and future leaders.
With Viesturs Liegis, Managing Partner, Amrop Baltics & Ukraine
The learning hidden inside junior work
Consider an executive search assignment. A researcher who maps a market does more than assemble names. They start to understand how an industry is structured, which companies develop particular kinds of leaders, and where a client’s assumptions may need to be challenged. Drafting a candidate assessment requires them to distinguish a convincing career story from evidence of performance. Discussing that assessment with a partner shows how facts, context and uncertainty come together in a recommendation.
AI can assist with every stage. That is valuable, especially when it frees people from repetitive tasks, but if the first analysis always arrives fully formed, a junior colleague may have fewer chances to frame the problem, notice what is missing and defend a conclusion. A polished document can make someone appear more capable than they would be if asked to explain the reasoning behind it.
This concern extends beyond executive search. Law, consulting, accounting and other advisory fields rely on experience accumulated through work that is often described as basic. If that work changes, the way firms develop expertise must change with it.
What the research tells us
A 2025 study published in the CHI Conference proceedings surveyed 319 knowledge workers about their use of generative AI in work tasks. The researchers found that greater confidence in AI was associated with less reported critical thinking, while greater confidence in one’s own ability was associated with more. They also observed a shift in where people applied their thinking: toward checking information, integrating responses and overseeing the task.
This was a survey of reported behavior, not proof that AI causes professional skills to decline. It does, however, reinforce a practical point: giving people access to a strong answer does not guarantee that they will understand it, question it or learn from it. Those capabilities need attention of their own.
Make development deliberate
Firms should identify the parts of a task that are valuable because they teach judgement, even when AI could complete them quickly. A young consultant could first write a short view on a market or candidate, then use AI to test it against alternative interpretations. In a case review, the senior consultant could ask what evidence would change that view. After a client meeting, the team could discuss why one risk mattered more than another.
These moments need not be elaborate. Regular comparison between a person’s initial reasoning, an AI-assisted version and the final client recommendation can make the learning visible. Simulations can give less experienced colleagues practice with difficult client conversations before they carry that responsibility alone. Exposure to real decisions matters too: it is difficult to learn judgement when you see only the finished presentation and never the discussion behind it.
The aim is to preserve the opportunity to develop the capabilities that routine tasks once taught. AI may even improve that process by offering immediate feedback and a wider range of scenarios. But someone must decide what a colleague should practice, and when efficiency should give way to learning.
Managers must make their judgement visible
Senior professionals often make decisions quickly because years of experience have taught them what to notice. To a junior colleague, that reasoning can be invisible. Managers therefore need to explain more about it: which signals they trusted, which they discounted, what alternatives they considered and what they still do not know.
This makes coaching a core part of the manager’s role. It also changes how we should recognize leadership. If a partner delivers excellent results but leaves the next generation unable to make sound decisions independently, the organization has a succession problem. Strong leaders build capability in others as well as delivering for clients themselves.
In our work at Amrop, this is relevant to how we think about leadership potential and succession. Individual performance remains important. So does the ability to coach people, share judgement and create conditions in which others can grow into responsibility.
AI offers professional services firms a real opportunity to improve how they work. The firms that benefit most over time will also ask what happens to the people learning the profession. If we automate parts of the apprenticeship, we must design the next one deliberately. The defining question is not simply how well our teams perform with AI today. It is whether we are developing the judgement they will need to lead tomorrow.
