introduction
Opinions on AI, and more specifically its impact on the labour market, are very diverse. There are those that make you dream (with the promise of unprecedented productivity and prosperity) but equally you read and hear the voices that prompt reflection (with a spectre of mass unemployment).
In this series of 5 contributions we choose not to put the technology at the centre, nor the macro impact on various societal domains, but instead focus on the way in which organisations and leaders can anticipate and facilitate the possible impact of this technology within their team, department, organisation. Because as with every far-reaching change: it is not the technology that determines whether a transition succeeds, but the human factor.
A first contribution brings together a number of voices in the debate that prompt a nuanced reflection.
yes but, it will all turn out fine
There is something reassuring about those who believe it will not come to that. "It is a hype that cannot be reconciled with the stubborn reality of real organisations and real work," they say.
In his Substack piece "An AInconvenient truth" Rutger Bregman points to the danger of that underestimation and describes how the "yes but …" keeps shifting along with the evolution of the technology. First it was "yes but, it is just a glorified autocomplete that guesses words", then "it can imitate, but it will never really reason", then "it passes exams, but no serious company actually uses it", then "companies use it, but the economics do not add up, it is one big bubble".
No one knows what the future holds, but the power of the technology is real. Perhaps it is worth looking to the past to learn something about the possible impact of this technology on work.
occupations versus tasks
The exercise of estimating the impact of a technological evolution on work was already carried out in earlier evolutions. In 2013 Carl Benedikt Frey and Michael Osborne published a study in which they assessed 702 occupations for susceptibility to automation. Their conclusion that 47% of American jobs were at high risk travelled the world. However, the study measured the technological potential, not the pace. Frey has since had to repeatedly clarify that he never claimed half of jobs would disappear within twenty years.
Three years later the OECD came up with a fundamentally different reading. By looking not at whole occupations but at the individual tasks within a job, they arrived at only 9% of jobs at acutely high risk.
The nature of the evolution in the case of AI is potentially of a different order of magnitude, but this 'occupations versus tasks' debate is still valid today. AI may rarely replace an entire role in one blow, but it potentially nibbles away at tasks and that leads to the reshuffling of roles. That sounds more reassuring, but it has a downside: it makes the uncertainty more prolonged, a creeping change rather than a shock.
impact or excuse
Recent figures seem to indicate that the impact of AI is now becoming visible in the statistics. According to the American Challenger report of May 2026, AI was the most cited reason for redundancies for the third month in a row, accounting for 38,379 announced jobs in a single month, the highest monthly total since this reason began to be tracked in 2023. Over the first 5 months of 2026 the counter already stands at 87,714 jobs, or 22% of all redundancies, more than in the whole of 2025.
And yet caution is advised with those figures. Economist Robin Brooks warns that the wave of redundancies among knowledge workers that is coming is not really about AI. Many of the jobs now being cut, he argues, lost their reason for existing long ago, often a surplus from the post-COVID hiring wave. For many companies, AI has become the socially acceptable excuse to restructure without bearing the blame.
a nuance for those looking from Europe
A caveat applies to all this: a Belgian or European leader cannot simply view the AI transition through an American lens, even though the loudest stories and the sharpest figures come from across the ocean. In the United States, 'employment at will' largely applies: dismissal can happen quickly and with few formal intermediate steps. That partly explains the Challenger figures and why AI is so readily given as a reason there.
In Europe it proceeds more slowly and with more structure. The EU AI Act contains protection for employees, and in countries such as Germany the deployment of AI in the workplace is closely tied to the co-determination rights of the works council. The German IAB estimates that over the next fifteen years 1.6 million jobs will be redesigned or cut, not as a shock, but as a gradual shift. Anyone in a European organisation who copies the American 'move fast and break things' logic risks not only legal problems, but above all a breach of trust that reverberates for years.
the human factor
Whether AI is the real cause or the story told about it, the effect on employees is the same: uncertainty, mistrust and a feeling of loss of control. It is not exactly the ideal mix for increasing the productivity of organisations.
If we assume that the impact of AI is coming, then the human factor will play an important role in the pace of the change and in achieving the anticipated benefits. It is not the technology that determines whether a transition succeeds, but the way in which organisations and leaders take the human transition phase seriously. The same technology can be a lever that strengthens people, or a crowbar that destroys trust.
At the bottom of this article you will find an overview of the other parts of this series.
We excel at guiding organisations through change, with a deep focus on the human factor. Feel free to get in touch whenever you'd like to exchange ideas.
