Congratulations, you’re a manager now

For generations, becoming a manager came with certain clues. There was a new title, a slightly better salary and, somewhere on the organisational chart, a handful of boxes suddenly appearing beneath yours. The people occupying them needed briefing, encouragement, correction and, from time to time, rescuing from an idea that had seemed perfectly sensible on Tuesday morning. The boxes are still appearing but they just aren’t necessarily occupied by people.

Boston Consulting Group’s latest AI at Work research, covering 11,749 workers across 14 markets, found that 47 per cent of respondents now spend more time managing and directing AI than doing the work itself. Among frontline white-collar employees without formal management responsibilities, regular AI use has reached 74 per cent. We hardly needed another survey to tell us that AI has entered the workplace; what is more interesting is the role it is quietly assigning to the people who use it.

Consider an increasingly ordinary working morning. You decide what to delegate to an AI system, brief it, provide context, inspect what comes back, discover that something delivered with impressive confidence is wrong, explain what needs changing and send it away for another attempt. Sometimes the result is excellent, sometimes mediocre and occasionally complete nonsense dressed in a very good suit.

This is beginning to look less like using software and more like managing work. Yet much of the corporate conversation still revolves around AI literacy, adoption and prompting, as though the central challenge were learning which buttons to press. IBM’s newly published global CHRO study suggests something more consequential is happening. Eighty per cent of CHROs say AI creates ‘invisible’ work for employees: validating recommendations, correcting mistakes, adding context and managing exceptions. When things go wrong, meanwhile, 43 per cent of employees say the blame still lands on them.

The machine may do the work, but accountability has not been automated.

There is considerable opportunity in this shift. Management has traditionally arrived relatively late in a career, after somebody has accumulated enough experience to be entrusted with other people’s work. AI scrambles that sequence: a young professional may now be delegating, reviewing and coordinating work long before anybody thinks of calling them a manager, turning capabilities once associated with seniority into part of everyday professional life. But there is a catch, and it deserves more attention than another workshop on clever prompts: how do you learn to manage work you have never learnt to do?

A young analyst once became useful by building the spreadsheet rather than merely inspecting it. A junior lawyer learnt by reading hundreds of documents; a researcher by following unpromising trails; an aspiring strategist by producing the first draft that returned covered in somebody else’s objections. Much of this work was repetitive and occasionally miserable, but it was also where patterns became visible, mistakes became memorable and professional instinct began to form.

If AI removes too much of that apprenticeship, organisations could eventually find themselves with a peculiar generation of managers: people highly practised at supervising output but less familiar with the craft required to produce it. Efficiency gained today could quietly become capability lost tomorrow.

This is where AI management becomes a question of reinvention and future relevance. Organisations will have to think beyond which tasks can be automated and decide which experiences people still need in order to develop. They will need to distinguish between work that is merely laborious and work whose difficulty is precisely what makes it valuable to learn.

For individuals, the shift is subtler but just as significant. Future relevance may increasingly depend on becoming comfortable orchestrating capabilities beyond our own while retaining enough knowledge of the underlying work to know when those capabilities are failing us. The organisation chart may continue to insist that nobody reports to you. Your working day is beginning to tell a rather different story: management has arrived but the promotion may take a little longer.