A survey conducted by the Pew Research Center revealed that job growth among men surpassed that of women in professions more susceptible to artificial intelligence (AI) in the United States. The analysis compared labor market data before and after the dissemination of generative AI to map gender disparities.
The study examined the evolution of the number of workers in the period between 2017 and 2019, and also between January 2023 and June 2026. The years 2020 to 2022 were omitted from the comparison due to changes caused by the Covid-19 pandemic in the work environment.
The findings indicate that the higher the degree of exposure of a function to AI, the more pronounced the difference becomes in job growth between men and women. Specifically, for every one-point increase on the AI exposure scale of an occupation, male employment registered a 6.3% increase, while the female variation was almost insignificant, with an advance of only 0.2%.
This discrepancy became even more noticeable in careers with high technological exposure. Male employment grew approximately 24% more in occupations most impacted by AI compared to those least exposed. In contrast, female employment growth remained practically stable across this exposure scale.
However, the scenario reversed in professions with less contact with artificial intelligence, where women showed higher job growth than men.
The survey emphasizes that the correlation between AI exposure and increased hiring does not necessarily imply that technology is the direct cause of these differences, given that other transformations occurred during the period and could have influenced the results.
To conduct this analysis, the Pew Research Center used data from 480 occupations, based on information provided by the Occupational Information Network (O*NET), a database maintained by the U.S. Department of Labor. Occupation classification takes into account the tasks performed in each profession, the feasibility of these tasks being assisted or performed by generative AI, and the relevance of these activities to work.
The evaluation scale used ranges from zero to 3.5; the higher the score, the greater the proportion of critical tasks in an occupation that can be supported or done by technology.
Professions with very low exposure include truck drivers, construction workers, and cleaning professionals. In the intermediate range are medical assistants, cashiers, and audiologists. On the other hand, occupations with very high exposure include accountants, software developers, and financial managers, as their activities involve information processing, reading, and writing, processes susceptible to assistance from generative AI.
The research also pointed out that the variation in job growth differs according to the educational level of the workers. Among individuals with a high school diploma or lower education, male employment grew by 8.8% for each additional point of AI exposure. For women at this same level of education, however, there was a reduction of 9.5%.
The difference was less pronounced among those who attended college but did not complete a higher education course; however, the gap between the results for men and women remained considerable in this group. Among workers with a bachelor's degree or higher education, no significant changes in job growth were observed for either gender as AI exposure increased.
Additionally, the study observed distinctions in the distribution of workers across professional categories. Men without a four-year university degree tend to be concentrated in activities less exposed to AI, such as cargo transport, construction, and merchandise handling. Women with the same level of education, on the other hand, more frequently work in roles such as receptionist, customer service, and secretarial work, occupations located in intermediate or higher positions on the AI exposure scale, suggesting a potential role for tools like chatbots and automated systems in these activities.
Finally, the Pew Research Center stresses that it is still premature to define the long-term impacts of artificial intelligence on the labor market, and the presented analysis focuses on the differences observed in job growth, without establishing a direct cause-and-effect link between technology adoption and changes in hiring.
