Employment growth differed for men and women across occupations with varying levels of AI exposure, according to a Pew Research Center analysis. The results raise questions about how labor-market shifts are distributed, but they do not show that AI caused the difference.
Pew reports that for each one-point increase in an occupation’s AI-exposure score, men’s employment grew by 6.3% and women’s by 0.2%. At the lowest AI-exposure levels, the pattern was different: women’s employment grew faster than men’s. These findings describe how employment growth varied across occupation groups; they do not identify what produced the variation.
How Pew made the comparison
The analysis compares employment in 2017–2019 with employment from January 2023 through June 2026, excluding 2020–2022. It groups employment by occupation and gender, then examines how growth relates to the occupations’ AI-exposure scores. Pew published the analysis on October 22.
That level of analysis matters. An occupation-level result describes a pattern across groups of jobs. It is not a tally of individual workers who lost or gained jobs, and it cannot establish what happened to any particular person. The results also do not show that individual men gained work because they used AI, or that individual women lost work because of it.
An AI-exposure score is part of the way the occupations are compared; the reported relationship does not by itself show that AI tools changed those jobs. The figures do not explain why employment growth differed by gender, either. The evidence therefore supports describing an association between occupational exposure levels and employment trends—not a causal account of those trends.
What the numbers do—and don’t—tell us
The 6.3% and 0.2% figures should be read in the context of Pew’s comparison. They describe the reported relationship between a one-point increase in an occupation’s exposure score and employment growth for men and women. They are not counts of jobs gained or lost, nor do they measure the employment outcome of an individual worker.
The lower-exposure result is another important part of the picture. Women’s employment grew faster than men’s in occupations with the lowest AI-exposure levels. So the analysis does not describe one uniform difference across all work: the reported pattern varies with occupational exposure.
Axios also covered the analysis on October 26 and noted that it does not establish causation. The central takeaway is a difference in employment growth across occupation groups that merits attention, not proof of an AI-driven shift. Pew’s analysis identifies the pattern, while leaving open what accounts for it.
