Anthropic research finds men use AI coding agents more than twice as often as women among social scientists

Anthropic studied how social scientists use AI and found researchers with typically male names use coding agents (AI tools that write program code automatically) more than twice as often as those with typically female names, with the gap holding even within the same disciplines and career levels.

Coding agent adoption varies sharply by field: economists lead at 39 percent, while education researchers are at 4 percent. PhD students and postdocs use coding AI far more than professors. The dominant use case is code generation for data analysis at 97 percent, while only a third use AI for writing text.

88 percent of respondents rate AI's effect on their own paper output above 5 on a 10-point scale, but 70 percent are more optimistic about their own productivity than about AI's impact on the social sciences as a whole. The authors suspect researchers worry that more papers could overload peer review and worsen existing problems like selective reporting.