Standing alone under nonstop fire

Picture an ordinary week for a manager in a large company right now.

HR has launched an AI assistant for leave and policy questions, with a learning path. Finance has a new expense agent, with a user guide. Procurement has moved intake to an AI front door, with an SOP. The practice wants everyone certified on two AI courses by quarter end. Accounts has a new copilot inside the ERP. Each one arrives with a launch email, a training module and a deadline.

None of these is a bad idea. Each team is doing its job. That is the problem.

Every builder sees one initiative

The people building each rollout see their own initiative, their own adoption curve, their own success metric. Nobody owns the view from the receiving end: one person absorbing six new ways of working in the same month while still running projects, teams and clients.

It feels like standing alone in a battlefield under fire from every direction. Not because any single shot is large, but because they never stop and they never coordinate.

Saturation is measurable

This is not just a feeling. In Prosci’s change management research, 73% of survey respondents said they were near, at or past the point of change saturation. And a Capterra finding that Prosci cites puts 54% of employees with change fatigue as looking for a new role.

Now add AI to that load. AI changes are not one change. They change the tool, the workflow, the skill expected and often the role itself, all at once.

What the rollout plan is missing

Most change plans are written per initiative. They ask whether this change is well communicated, well trained and well supported. They rarely ask the question that decides whether any of it sticks: how much change is this person already carrying?

Three things would help, and none of them needs new technology:

1. One calendar for change. Someone owns the combined load on each role, the way someone owns the release calendar in engineering. Launches get sequenced, not stacked.
2. A budget for attention. Treat employee attention as a finite resource with a quarterly limit. A new rollout has to fit, or something else waits.
3. Measure use, not completion. Course completions and certifications count activity. What matters is whether people work differently a month later.

Try it yourself

I built The Pressure Line to make this visible. Load a team’s quarter with initiatives, set how disruptive they are and how much real support people get, and watch the gauge. No single change breaks it. The stack does.

AI will scale in an organisation only as fast as the people in it can absorb it. The rollout plans that forget this are not moving fast. They are moving the cost onto the people least able to push back.