Hire for the org chart AI is drawing

Open most senior job descriptions in a data or AI-adjacent function and you’re reading a careful description of how the role works today: the team it manages, the process it owns, the deliverables it’s accountable for. The trouble is that AI is redrawing exactly those things while the req sits in the market, and, as we’ve argued in how AI changes the tasks before the titles, it comes for the work long before the label admits it. The team that role manages is about to be smaller and more senior. The process it owns is about to be half-automated. You’re hiring against a snapshot of a role that will look materially different by the time the person is a year in.

This is a warning that the most confident, most specific reqs, the ones that read as rigorous because they nail today’s role in detail, are often the most brittle. They select precisely for fitness to a job that’s dissolving.

The redraw isn’t cosmetic. AI takes the work at the bottom of a function first: the repeatable analysis, the first-draft build, the routine review pass. When that work leaves, the shape of the team above it changes with it. A leader who used to run 12 people executing known processes ends up steering 5 who each carry more judgement and own more of the outcome. The pyramid flattens and rises at the same time. That’s not a shape I’m inventing to make a point; the World Economic Forum’s Future of Jobs 2025 puts 39% of core skills changing by 2030. When the skills move that fast beneath a function, the org that sits on top of them can’t hold still either.

So the seat you’re hiring for is quietly shedding the parts that made it legible. Less headcount to coordinate, more ambiguity to resolve, and more of the value in deciding what the smaller team should build at all. The req still describes the old pyramid, because that’s the org that exists on the day someone writes it. The one your hire will actually run is already forming underneath it.

The market’s reflex is to future-proof by bolting AI keywords onto the req: hire someone who’s ‘used the tools,’ who lists the right frameworks. That’s hiring for today’s tooling, which will be obsolete faster than the role. The thing actually worth selecting for is a disposition: the ability to lead a function whose definition keeps shifting, to redesign the work as the tools change, to hold judgement steady while the mechanics move.

That leader looks different on paper. They may not have the deepest command of the current stack, because they’ve spent their career adapting to three previous ones. What they have is a track record of walking into a role, watching it change under them, and reshaping it rather than defending its old form. In a moving org chart, that’s worth more than mastery of the version that’s about to expire.

The fair worry is that ‘adaptability’ is the softest thing to hire for, a word every candidate has learned to perform. But the trait shows up in specifics once you ask for them. I want to hear about a real role that changed shape while they held it: what they stopped doing, what they took on, how they rebuilt the team around the new reality. A leader who has genuinely lived through that tells it plainly, with the awkward details intact. The one who hasn’t reaches for the vocabulary and stalls the moment you ask for the story.

The clearest tell is how they talk about the current stack. If a candidate’s whole edge is command of today’s tools, they’re anchored to the version of the job that’s expiring. The leaders built for a moving chart describe their edge as command of the problem, and treat the tools as things they expect to swap out twice more before they’re done. That difference sounds subtle across a table. It decides whether the person is still the right hire once the tooling turns over, which in an AI-touched function is a question of quarters.

The practical discipline is to draft the role as you honestly expect it to be a year and a half out: smaller team, more impact per head, more ambiguity, more of the leader’s value in framing problems rather than running known processes. Then hire against that. It’s uncomfortable, because it will disqualify candidates who are perfect for today’s version and unremarkable against tomorrow’s. That discomfort is the signal you’re doing it right.

Hiring for the org chart you have feels safe and precise. It just quietly locks your leadership bench to a structure AI is dismantling. The centres that will still have the right leaders in place when the dust settles are the ones drawing the future chart now, the one our three-year outlook for GCC hiring keeps pointing at, and staffing that one.