Where AI hits the wallet.
The fear says AI comes for the good jobs: graduate, white-collar, high-paid. £288bn of UK wages do sit in high-exposure (AI 7+) roles. This page follows the money: which pools sit where on the exposure curve, why the worst-placed money is in the middle, and the one place the bite has a gender.
The money isn't all at the high end.
The biggest wage pools sit at exposure 3 and 7. The worst-placed money is the unclear middle (5–6): £188bn of wages on the lowest average outlook in the dataset.
The one place it bites hardest has a gender.
Same AI exposure (7+), opposite directions, and the workforces are mirror images. The roles AI is set to shrink are 69% women; the ones it's set to grow are 68% men. AI doesn't close the occupational gender gap. It locks it in.
The full gender cut: including the measured twist that reverses the usual claim →
The per-occupation wage pools come from the 738-occupation dataset (NCS pay × ONS employment); they are bases in play, not forecasts of wages lost. Median UK earner: HMRC / ONS all-earner income percentiles.