Eighty-three points separate the most and least protected jobs here, and the biggest occupation in the group is one of the most reachable in the whole economy.
Exposure means published research judged an AI could halve the time a task takes, at the same quality. It is not a prediction about any job. How to read this
That is the highest "today" share of any occupation of that size in the report. More than half the listed work can be halved with a chat window and nothing else.
It is also the occupation most people picture when they imagine AI taking jobs, so it is worth being precise: the rating says the work gets faster, not that the role ends. What happens next is a decision employers make, and no dataset here predicts it.
Legal Secretaries and Administrative Assistants are the least protected in the group. At the other end, Mail Clerks and Mail Machine Operators are at 83 per cent protected, because the work involves physically handling things.
Same group, same label, opposite ends of the report. The word "admin" is doing no useful work as a category.
Solid is reachable with a chat window today, pale needs software building first, grey cannot be sped up. Task counts are shown because several of these rest on fewer than twenty.
Two of the nine People occupations sit in this group, and both are among the roles most often described as entry-level or a stepping stone.
The chapter is worth reading before anybody repeats that. The group contains the most reachable large occupation in the economy and one that is 83 per cent protected, and the difference between them is not seniority.
There is no data here on what anybody actually does with AI. A dataset that appears to measure it was removed from this report because its meaning could not be stated. The reasoning is on the limits page.
The exposure ratings cover all 1,012 tasks in these 51 occupations with no gaps, and were made in 2023 against the model capabilities of that year. Read every figure as a floor.
Exposure ratings from Eloundou, Manning, Mishkin and Rock, GPTs are GPTs: Labor market impact potential of LLMs, Science 384, 1306–1308, 2024, used under the MIT licence. This page includes information from the O*NET Database by the U.S. Department of Labor, Employment and Training Administration (USDOL/ETA), used under the CC BY 4.0 license. O*NET® is a trademark of USDOL/ETA. People Team AI has modified all or some of this information. USDOL/ETA has not approved, endorsed, or tested these modifications. Employment figures from the US Bureau of Labor Statistics.