Will AI replace junior roles in professional services?

What work are juniors actually doing today?

Walk through a junior fee earner's week in most professional services firms and you will find a lot of assembly. Pulling documents together for a bundle or a file. Writing up notes from a meeting a partner led. Drafting the standard sections of a report so a senior colleague can write the judgement parts. Chasing figures, reconciling versions, reformatting, checking names and dates. Producing the first draft that someone more senior will then rewrite.

That is exactly the work AI is best placed to take over. It is structured, it repeats, and it draws on information the firm already holds. So the fear is understandable. If a rebuilt process can produce the first draft of a report or the write-up of a meeting in minutes, what is the associate for?

Is the fear justified?

Partly, and it is worth being honest about which part. Some tasks that have filled junior timesheets for a long time will shrink sharply. A firm that has built its model around large numbers of juniors doing high volumes of assembly will find that model under pressure, not least from clients who have always been reluctant to pay for it.

But replacement is the wrong way to think about it. Firms do not employ juniors only to produce first drafts. They employ them because today's associate is the next decade's senior fee earner and partner. That pipeline does not stop mattering because drafting got faster. If anything, it matters more, because judgement becomes the thing the firm sells, and judgement has to be grown.

The firms that treat AI as a reason to stop hiring juniors are making a short-term saving against a long-term problem. In a few years they will have partners who are stretched, and nobody below them ready to step up.

What is the real risk, then?

The real risk is quieter. For generations, professional training has worked through doing. A trainee learns what a good report looks like by drafting dozens of bad ones and having them corrected. They learn what matters on a file by assembling files and seeing what the partner asks about. The dull work was also the apprenticeship.

Take the dull work away and the apprenticeship goes with it, unless something is put in its place. A junior who has never drafted a section from scratch will struggle to review one properly. A reviewer who does not know how a document is built will miss the errors that come from how it was built. And a firm that keeps a named person accountable for everything that goes to a client needs people who can genuinely do that job, not merely click approve.

This is the part of the change we think firms are least prepared for. The technology question gets the attention. The training question gets almost none.

What should junior roles become?

The work does not disappear. It moves up a level. In a firm where assembly is handled by a rebuilt process, a junior's time can shift towards things that develop them faster:

• Checking, not just producing. Reviewing a complete draft against the source records, questioning figures that look odd and tracing each statement back to where it came from. That builds the habits a senior reviewer needs.
• Earlier contact with clients. With less time spent on assembly, juniors can sit in on more conversations, take more calls and handle more of the routine updates, with support.
• Owning part of the process. Someone has to notice when the output drifts, when a template needs changing or when a new kind of job does not fit. That is a good job for a sharp junior and it teaches them how the firm's work fits together.
• Deliberate drafting practice. Some firms will choose to have trainees draft certain documents by hand, for learning rather than for delivery, and compare their version with the finished one. That is a conscious choice, not a return to the old way.

None of this happens by default. It happens because someone senior decides what the first two years should teach and redesigns the work to teach it.

What about the partners?

There is a connection here that firms often miss. Juniors today spend much of their time producing drafts that partners then rewrite, which is a large part of why partners become the constraint on the whole firm. We make that argument in the partner bottleneck. When drafts arrive closer to finished, partners have time to supervise and teach properly, which is the one thing the old model rarely left room for. Fixing the process helps both ends of the firm at once.

How should a firm approach this now?

Start with one process and look at it through a junior's eyes as well as a partner's. Before it is rebuilt, ask who currently does each step, what they learn from it and what they will do instead. After it goes live, check whether the juniors on that work are developing faster or simply being given more volume. Meeting and file notes are a good place to test this, because they are a classic junior task and a natural source of learning about a matter. The file notes and case notes page describes that job.

Be open with the team about what is changing and why. People early in their careers read uncertainty as threat, and a clear account of their new role does more for adoption than any amount of training on the tools. Our guide on getting a team to adopt AI covers that side. If the question in the room is really whether to hire or automate, automation or hiring sets out the trade-offs. The honest answer is usually both, with the hiring aimed at people the firm intends to develop, not at more hands for assembly.