The objection you will hear

At some point in almost every AI capability programme, a business manager says some version of the same thing: 'I support the training, but I cannot give you two days right now. We have a quarter to close.' The L&D director nods, the dates shift, and the programme quietly loses the people who most need to be in the room.

The problem is not that the manager is wrong about the quarter. The problem is that the conversation is framed as a competition between learning time and productive time. Once it is framed that way, productive time wins. Your job is to reframe it before that conversation starts.

Why the framing matters

Most AI capability programmes are still designed around the logic of a conventional technology rollout: the organisation chooses a platform, schedules training, and expects adoption to follow. That sequence made sense when software was stable and the learning task was bounded. Generative AI does not work that way. The tools change quickly, the useful applications vary by role and context, and the gap between knowing about AI and being able to use it on real work is wider than a half-day awareness session can close.

The result is that many employees have already started without the organisation. They are using personal accounts, experimenting privately, and building habits that may or may not align with how the organisation wants the work done. My article on the AI adoption that already happened describes this pattern directly: the organisation is often discussing how to introduce AI to a group that has already introduced it to itself. The adoption that matters has already begun. The question is whether it happens in the open, where it can be improved and governed, or in the background, where it cannot.

That changes the argument you bring to a business manager. You are not asking for time to introduce something new. You are asking for time to make current practice safer, more consistent and more useful to the team.

What practice time actually needs to do

In the programmes I design, my delivery is around eighty-five percent experiential: clinics, workshops, applied work and simulation. The aim is for participants to leave having practised the work they will be expected to do next. The Guided Mastery method that underpins this approach rests on three principles: incremental challenge, learning by doing and immediate application. Participants leave having practised the work, not merely heard about it.

That distinction matters because hearing about AI and doing AI on a real task are separated by a gap that good intentions do not close. Practice time therefore needs to be structured around actual work, not illustrative exercises. A prompt-writing drill using a fictional brief is less useful than a prompt-writing session using the brief that is due on Friday. The learning transfers because the context is real. The manager's concern about lost productivity is also reduced, because the output of the session is something the team actually needed.

This is the practical argument: structured AI practice on live work is not time taken away from the work. It is time spent on the work, with a learning frame around it.

A hypothetical planning example

Consider a hypothetical L&D director at a professional services firm preparing a capability programme for a group of senior analysts. The business unit head is supportive in principle but has said she can have ninety minutes per week, not the half-day blocks she originally proposed.

Rather than negotiating for more time in the abstract, she maps the team's actual recurring tasks: client briefing summaries, competitive landscape reviews and internal status reports. She identifies which of those tasks are good candidates for AI-assisted drafting, and she designs three short practice sessions, each built around one of those tasks.

In the first session, analysts use the organisation's approved AI environment to draft a briefing summary from a set of source documents. The output is reviewed, the prompts are refined, and the team discusses where the draft was useful and where it needed correction. The session takes eighty minutes. The team leaves with a draft they can actually use and a clearer sense of where human judgement is still doing the essential work.

The business unit head sees a completed briefing summary come out of the session. The next conversation about time is easier.

This is hypothetical, but the structure is transferable. The key moves are: anchor the session to a real deliverable, make the output visible to the manager, and keep the session short enough that the time cost feels proportionate to what comes back.

The tradeoffs to name explicitly

Being honest about tradeoffs strengthens the case rather than weakening it. There are three worth naming before a manager raises them.

  • Short sessions on real work produce narrower learning than longer structured programmes. Ninety minutes on a briefing summary builds one skill in one context. It does not build the broader judgement about when to use AI, when to question its output, or how to assess a use case. A programme that only ever runs in short bursts will produce capable users of one workflow and leave the harder thinking undone. The L&D director needs to be clear about what the short sessions are building toward, not just what they deliver in the moment.
  • Anchoring practice to current tasks can entrench current workflows. If analysts only ever practise AI on the briefing summaries they already write, they may become faster at producing briefing summaries without ever asking whether the briefing summary is the right output. Some practice time should be reserved for tasks that are slightly outside the current routine, so that the learning includes the question of what the work could become, not only how to do the existing work more efficiently.
  • Visible outputs create pressure to optimise for the output rather than the learning. If the business unit head starts treating practice sessions as a production resource, the learning frame disappears. The L&D director needs to protect the review and reflection element of each session, even when it feels slower than simply finishing the task.

Naming these tradeoffs in the planning conversation signals that you have thought past the pitch. It also gives the manager something to hold you accountable to, which tends to make the agreement more durable.

What the negotiation actually requires

The practical ask is not a block of time. It is a decision about who owns the learning within the work. Business managers own the output. L&D owns the structure that makes the output a learning event. Those two things can coexist in the same ninety minutes if the roles are clear before the session starts.

The measure of success is not attendance or session ratings. It is whether the team does something differently the following week, and whether the manager can see it. That is the outcome worth tracking, and it is the outcome that makes the next conversation about time considerably shorter.

Written for this site.