Most organisations have done the AI awareness training. Their people know what ChatGPT is. Some use it occasionally. Nothing has changed about how the work actually gets done.

Buying the platform is the easy part, and it is the part most organisations have already finished. The competence is what nobody budgeted for.

Competence before platform

The usual sequence is backwards. Licences are issued, an announcement goes out, a training session is run, and then leadership waits for productivity that does not arrive. The tool was never the constraint.

What changes the number is people becoming genuinely good at a small set of things that matter in their actual role, coached over weeks rather than briefed in an afternoon.

Where this comes from

I coach Microsoft 365 Copilot adoption inside a large South African insurer, as one of a five-facilitator team, week over week across multiple cohorts. That programme has been running since June 2026.

So the material is not a survey of what other people have written about AI adoption. It is what happens when the same cohort comes back next week and tells you what did and did not work in their real job.

What the talk covers

• The three reasons adoption stalls, and why only one of them is technical
• What competence looks like when it is measured rather than assumed
• Why the people who adopt fastest are rarely the ones you expect
• How to tell the difference between a team using AI and a team performing use of AI
• What leadership has to do differently, which is more than sponsoring the licence

Format

• Keynote, 45 to 60 minutes, for conferences and leadership forums
• Workshop, half or full day, where the team works on their own tasks
• Both can be tailored once I know what tooling you have already bought