Careermash
EDITION ONE · MEASURED AUGUST 2026

State of the professions

1,840 UK careers, measured against what AI is already used for. This is the Careermash Exposure Index, edition one: computed live from the same engine as every card, and from edition two it tracks what moved. Nothing here is a prediction.

The distribution

Median exposure across all scored careers is 19; the mean is 23. The spread is the story: most work still measures low, and a hard tail measures very high.

0-241,061 careersAI is a marginal presence in the measured tasks
25-49627 careersAI is established in a meaningful share of the work
50-74152 careersAI is doing much of the measurable work already
75-1000 careersThe measured tasks are largely open to AI today

The moats

Of the 1,512 careers with a classified moat:

431Physical (hands-on work)
407Regulatory (protected by law)
259Relational (people want a real person)
415None (no special protection)

328 careers hold a real exposure score but no moat classification yet; their cards say so honestly rather than guessing. The register grows only through reviewed evidence, and every change lands in the public ledger.

Ties are shown as ties: within a shared score the order is alphabetical because the measurement supports no finer ranking. Full data: open data · careers.json.