ChatGPT Adoption in Academia Highlights a Stark Gender Productivity Gap

After the launch of ChatGPT, male researchers' productivity rate increased, pointing to a gender difference in how scientists use AI. This finding is based on preprint uploads to the Social Science Research Network (SSRN), a major open-access

​​​​​​​Research: Gender disparities in the impact of generative artificial intelligence: Evidence from academia. ​​​​​​​Image Credit: bombermoon / Shutterstock

After the launch of ChatGPT, male researchers' productivity rate increased, pointing to a gender difference in how scientists use AI. This finding is based on preprint uploads to the Social Science Research Network (SSRN), a major open-access repository, rather than all academic publications. Previous work has shown that men are more likely to use generative AI in their work than women. Shaobo Li and colleagues analyzed preprints uploaded to the SSRN from May 2022 to June 2023, applying a difference-in-differences (DiD) approach to estimate the impact of ChatGPT on research productivity. ChatGPT was released in late November 2022. After the release of ChatGPT, the productivity of male researchers significantly increased relative to that of female researchers, with the probability of them posting a preprint each month rising by 6.4%, a statistically significant difference.

Countries where ChatGPT has higher penetration showed a more pronounced gender disparity than countries where ChatGPT is seldom used. The study confirmed this using a difference-in-differences-in-differences (DDD) analysis, finding that the effect was strongest in countries such as the U.S., Spain, and Australia. In a separate survey of 400 researchers reached through Qualtrics, female researchers, on average, spent less time using AI and were less likely to perceive the tool as improving their efficiency than male researchers. This difference in perceived efficiency was largely explained by usage frequency—men were more likely to engage with AI tools, and this higher engagement accounted for their greater efficiency gains.

According to the authors, generative AI exacerbates gender-based productivity disparities, and institutions should factor such disparities into their faculty evaluations. While male researchers became more productive, the study found that the quality of their research—as measured by preprint downloads and abstract views—remained unchanged. The authors also highlight the need for more training opportunities for researchers to learn to use generative AI. Gender disparities in research productivity are shaped by multiple factors, including differing service burdens in academia and potential lingering effects of the COVID-19 pandemic, which the study controlled for where possible.