Why AI Training Doesn't Stick

A round table on adaptive and blended learning.

Our own research found a gap that is hard to look away from. Sixty-nine percent of organisations want to invest in AI skills. Twenty percent of people actually take part in training. Something breaks between the budget and the classroom, and most of us know it.

Part of the answer is that corporate AI training is built for a room that does not exist. One course, one pace, one starting level, and a group of people who are nowhere near each other. Adaptive and blended learning promise a way out: the programme adjusts to the learner, online and in person are mixed, and the classroom time goes to what genuinely needs a teacher.

That is the theory. In practice it raises harder questions. What do you measure? What do you do with that data? And what still falls apart without a human in the room?

In July we sat down with hiring managers and L&D professionals to ask which AI skills actually matter. This round table takes the next question: why people do not end up learning them, and what works instead.

This is also where our next annual research project starts. The 2027 edition of The Skills of Tomorrow’s AI Workforce is about AI adoption, and this conversation feeds directly into it.

No presentations and no pitch. A small group, a facilitated conversation, and a short write-up afterwards for everyone who attends.

For who L&D leads, heads of people development, and hiring managers responsible for AI and data capability in their organisation.

Practical Thursday 26 November 2026, 14:00 to 16:00 Matrix ONE, Science Park 301, Amsterdam Free of charge. Places are limited.

Request your place below.