For weeks we asked why no one running the country would name what's happening to work. This week the man about to run it did — and his own industry says the plan doesn't add up.

Open your phone this morning and it is all one story: missiles in the Gulf, oil spiking, the market down. It is a serious morning, and our desks cover it below. But the story that will still matter next month is quieter, and closer to home. Tomorrow, 9 July, nominations open in the contest to replace Keir Starmer, and unless someone forces a fight, Andy Burnham is Labour leader by the 17th and Prime Minister soon after.

Give him this much. For the first time in this whole saga, the person most likely to hold Number 10 has a team putting the future-of-work question on paper. The Financial Times reported on 2 July that his advisers are drafting an AI strategy built around a phrase they have been testing — "making tech work for people": retraining for workers whose jobs are exposed to automation, limits on who owns the data centres going up across Britain, more muscle for the competition regulator. After weeks of leadership pitches that left the machines out entirely, someone finally wrote them in.

Then read past the slogan, and it does not hold together yet. His own sector says so. Founders and analysts spent the week asking whether this is a programme or a list of grievances. One adviser line — questioning the "point" of driverless cars and "who they are for" — was meant, I think, to show concern for the drivers about to be automated out of a living. Instead it made people wonder who is advising him. Naming a problem is not the same as answering it. We have said for months that Westminster does not understand what is coming for work. The first serious attempt to prove us wrong is a set of gestures that has not been costed, sequenced, or squared with itself.

I am not going to tell you tech sovereignty is right or the old US-friendly model was wrong. Both camps are making big claims and neither has earned the benefit of the doubt: the outgoing approach was sold as prosperity and has not obviously delivered it; the incoming one is sold as protection and is not yet a plan. That is not fence-sitting. It is refusing to hand you a verdict the people proposing this have not done the work to earn.

One thing I owe you plainly, high up, because it is the sharpest conflict this paper carries. This note is mine — the view, the argument, the call to run it — and it was drafted with the same AI that writes the rest of the paper. And the plan above is a direct answer to the month Washington briefly switched off the American AI models that British firms, and this newsroom, had built their work on. Every analyst across these pages runs on models built by Anthropic, one of the companies at the centre of that episode. We report on this industry from inside it, and we tell you so.

So welcome the naming. It is more than we have had from anyone else on that bench. Just do not mistake it for the plan. A decade of change is already arriving. Tomorrow we start handing the country to a man who has, at last, noticed — and has not yet worked out what to do about it.

— M.