I have been asking myself a version of this question for two years now. The pace of AI progress became impossible to ignore. Not whether it matters. Whether the story we keep telling about it is actually the right one.
The story goes: AI is replacing jobs, wholesale, sector by sector, and the only question left is how fast. I do not think that is what is actually happening, at least not yet. Most organisations I see have not deployed AI into their processes by firing entire teams. What they have done, unevenly, is start asking different questions about what a role is for.
That distinction matters, because it changes what kind of change AI actually is.
I read a piece from Talsom this week on where change management is heading in 2026. One line stood out more than the rest. AI is not just changing processes, it is changing the psychological contract at work. Roles shift, decisions move, and the sense of what makes a person’s contribution valuable gets renegotiated. Often without anyone naming that renegotiation out loud.
That is the part almost nobody talks about honestly. Not the job losses, real as they are in specific sectors. The less visible thing: people staying in the same seat, doing recognisably the same job title, while what actually earns them their place has shifted under them without an announcement. In change management terms, that is a break nobody named. And an unnamed break does not stop people from resisting it. It just makes the resistance harder to read, because nobody can point to the moment it started.
There is a second mutation running in parallel that I think gets separated from this one when it should not be. Climate. Every industry is also being asked, at the same time, to account for what its way of working actually costs the planet it operates in. Two major, structural, non-optional changes, landing on the same organisations, in the same years, often handled by two completely different teams who barely talk to each other.
I do not think that is a coincidence worth ignoring. Both are the same kind of pressure: the ground an organisation has stood on for years is no longer sustaining it in the way it used to, whether that ground is a stable definition of a role or a stable assumption about resources. Neither is optional to sit out. Both ask the same question underneath: what do you actually carry forward, and what do you finally have to let go of.
My honest opinion, not the tidy one: AI is not one change, it is at least two, tangled together. A technical one, which is genuinely uneven and slower than the headlines suggest. And a psychological one, which is faster, harder to see, and already well underway in rooms that have not deployed a single new tool yet. Change management that only prepares people for the first one is going to keep being surprised by the second.
Source: Talsom, “Gestion du changement en 2026 : cinq leviers clés pour réussir dans un contexte en mutation accélérée.”
If you wish to be supported during your transformation projects, we invite you to schedule a meeting with The Change Society Office.
