AI transformation is usually described as a list of things to start. But many organisations stall because they never stop anything. They add AI on top of old processes, old reports and old assumptions, then wonder why nothing improves. Here are ten things leaders should consider stopping to make room for AI to deliver.
Processes to stop
1. Stop producing work nobody reads
AI makes it cheap to produce reports, which can multiply waste. Before automating a report, ask whether anyone uses it. Often the right move is to stop it altogether.
2. Stop duplicating approval steps
Work that passes through several approvers was often designed for a slower world. If AI speeds up drafting but approval still takes a week, nothing changes. Review whether every step is still needed.
3. Stop manual data re-entry
Copying information from one system or document to another is exactly the kind of work AI and simple automation can remove.
4. Stop starting every document from blank
Use templates, previous examples and AI drafts as starting points. A blank page is rarely the best place to begin.
Habits to stop
5. Stop treating AI as an IT project
AI changes how work is done. Process owners must lead it, with IT supporting.
6. Stop running pilots without end dates
Every pilot needs a decision date. Open-ended experiments drain resources and goodwill.
7. Stop measuring effort instead of outcomes
Hours spent and documents produced matter less when AI changes the effort involved. Focus on quality, speed and results.
Assumptions to stop
8. Stop assuming senior people cannot learn AI
Senior leaders often learn AI quickly once they try it on their own work, and their involvement accelerates everything else.
9. Stop assuming AI output is either perfect or useless
AI output is usually a useful draft that needs judgement. Teams that expect perfection give up; teams that dismiss it miss real gains.
10. Stop assuming you need perfect data or a perfect strategy first
Most organisations can start with general AI tools and a one-page direction, then improve as they learn.
How to decide what to stop
Run a short "stop review" with each team as part of any AI redesign:
- Which tasks, reports or steps exist mainly out of habit?
- Which would we not create if we started today?
- Which does AI make unnecessary rather than just faster?
Stopping work is often the fastest win in an AI transformation, because it saves time immediately and costs nothing.
The leadership challenge
Stopping things requires authority and courage, because every process has someone who feels attached to it. That is a leadership job. Our Leading AI Innovation & Organisational Transformation course includes redesigning your own workflows to save real time, which often starts with deciding what to drop. For related thinking, see AI hype vs real business value.
Frequently asked questions
Is it risky to stop processes during AI transformation?
Test first: pause a report or step for a month and see if anyone notices. If nobody does, stop it permanently.
Who should decide what to stop?
The team doing the work, with their manager's approval. They know which steps add value and which are habit.
Should we automate a process or eliminate it?
Always ask whether it should exist at all before automating. Automating unnecessary work just makes waste faster.