Jobs without a degree most exposed to AI
O*NET Job Zone 1 or 2, which is little or some preparation and no four year degree, ranked from the highest exposed share downward. Release v2026.Q3, capability reference date 2026-09-15, built on O*NET 31.0.
The usual story is that AI comes for graduate work first. For desk jobs that never needed a degree, much of the work is already producible, and there is less of a credential standing in the way.
How to read this
Exposed is the share of an occupation’s weighted task load that a generally available AI system can produce today with little structural friction. Assisted is work it can produce but cannot finish alone, because somebody has to be present, accountable, or because checking the output costs as much as doing the work. Untouched is work it cannot produce at all.
Every figure here is computed from task level ratings, and every occupation links to the page that shows its own tasks. If a rating looks wrong, the task is the level to argue at: how this is measured.
Questions about this list
What tops the list?
Telemarketers, with 73.3% of its weighted task load exposed to current AI systems, 22.1% assisted and 4.7% untouched. Release v2026.Q3.
How is this list decided?
O*NET Job Zone 1 or 2, which is little or some preparation and no four year degree, ranked from the highest exposed share downward. Scores come from rating every standardised work activity in each occupation against a published rubric, not from rating the occupation as a whole.
Does a high score mean the job disappears?
No. Exposure is not displacement. This measures what current systems can produce and what structurally stands in the way, not what an employer will do about it.
Other cuts of the same data
Exposure is not displacement. This index measures what current systems can produce and what stands in the way. It does not forecast job losses, and it models no adoption, employment or wage effect.