AI isn’t taking jobs. Employers are pulling up the ladder.

Stanford economist Erik Brynjolfsson analysed the payroll records of millions of American workers. His conclusion: since the breakthrough of generative AI in late 2022, employment for 22- to 25-year-olds in AI-exposed occupations has fallen by around 13 per cent. Junior software developers, accountants, customer service staff. And the experienced professionals in the same occupations? Stable. Or even growing.

Read that sentence again. AI isn’t replacing people. AI is replacing beginners.

And now it gets painful. The very employers sawing off the bottom rung of the ladder are complaining about a shortage of skills. In the Future of Jobs Report, around 63 per cent of employers cite the skills gap as the biggest barrier. And McKinsey put its finger on the sore spot: nearly half of employees name training as the most important condition for using AI well, and nearly half receive virtually no training.

So no, it is not the young people. Anyone claiming that the generation of 25 to 35 has ‘insufficient experience with AI’ has it backwards. This generation uses AI constantly in private. What it lacks is context: how do you deploy these systems in real decision-making, with real clients and real risks? You do not learn that from a prompting course. You learn it on the job. Precisely the job that starters find ever harder to land.

The doom-mongers are getting pushback from an unexpected corner. In A Century of Plenty (January 2026), Sven Smit, McKinsey partner, based in Amsterdam of all places, and his colleagues at the McKinsey Global Institute argue that AI is not a job destroyer but the greatest productivity engine in a century. Their sums: generative AI could add 0.5 to 3.4 percentage points to annual productivity growth. More than the steam engine. More than the IT revolution. And that productivity is a dire necessity, because in two thirds of all countries the birth rate sits below replacement level. We are simply running out of hands. But beware: those gains are not evenly shared. In a sample of 8,300 large firms, two thirds of productivity growth came from fewer than a hundred standouts, firms that keep investing through uncertainty, including in their people. Growth is not inevitable; growth is built. Optimism is not a strategy. Training is.

A handful of organisations already get it. IBM, recently featured in the Financial Times’ Working It newsletter, had taught 22 million people free AI skills through SkillsBuild by the end of 2025. Ikea rolled out AI literacy to tens of thousands of employees. And Moderna learned the most important lesson with its AI Academy: one size fits all does not work. One generic course became six learning tracks, embedded in the context of people’s actual work. That is the difference between a tick in a box and a skill.

Why should this concern you as a leader? Because your middle manager of 2031 is applying for a job right now. The thirty-somethings who will have to run the complex parts of your organisation, and the forty-somethings who must lead it after that, come from precisely the ranks now being thinned out. Whoever chokes off the intake today will be buying in leadership externally in five years’ time. At premium prices. Believe me, I know where those prices sit, it is my trade.

Three things you can do tomorrow:

1.    Keep hiring starters. Not out of charity, but out of enlightened self-interest. Redesign the junior role: less routine work, more learning with AI.

2.    Train in the context of the work. A two-hour e-learning module is theatre. Let teams practise on real cases and share what works.

3.    Make your thirty-somethings the owners. It is precisely the generation with some work experience that is most comfortable with AI. Give them the role of champion, it doubles as their apprenticeship in leadership.

The technology is not the problem. Nor are the numbers. The only thing missing is leadership that looks beyond the next quarter.

You are warmly welcome to exchange thoughts on finding and selecting the right people around you. Or for coaching conversations on developing yourself or those around you.

Warm regards,

Aegeus