Business and finance runs the full range: one occupation has no protected work at all, another has 100 per cent. A single figure for this group is close to meaningless, which is what makes the spread the story.
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
Financial Quantitative Analysts have no tasks rated beyond an AI's reach. Farm Labor Contractors, filed in the same group because both are business occupations, have 100 per cent.
Nothing about "business and finance" as a category predicts where a job sits. What predicts it is whether the work happens on a screen or in a room, which is the finding that holds across the whole report.
Accountants and Auditors employ 1.32 million people, and 87 per cent of their listed tasks are reachable in some form: 17 per cent with a chat window today, 70 per cent with a tool built on top.
That 70 per cent is worth staring at. It is the second largest middle band of any occupation in this chapter, and accounting software is an old, crowded, well-funded market. If the tools were going to appear anywhere first, it would be here.
Human Resources Specialists, Training and Development Specialists, Compensation and Benefits Specialists and Labor Relations Specialists all sit in this group. So do the analysts and the auditors they work alongside.
The People occupations are not outliers here. They sit in the middle of a group whose defining feature is that the label predicts nothing.
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.
Four of the nine People occupations live in this group, and their profile is the group's profile: a small reachable slice, a very large middle band, and a thin protected core.
The useful comparison is Accountants and Auditors. Same shape, same size of middle band, and a software market that has been building against it for forty years. That is what a mature version of the People profession's backlog looks like.
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 1,008 tasks in these 48 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.