Gender Diversity in AI Research: Key Report Findings

July 17, 2019

Cynthia RudinHow gender diverse is the workforce of AI research? Key findings from a large-scale analysis report recently published by Nesta UK (previously known as National Endowment for Science, Technology and the Arts) indicate:

  • A serious gender diversity crisis exists in AI research. This new report shows that less than 14% of AI papers feature a female co-author, and the proportion of AI papers co-authored by at least one woman has not improved since the 1990s. Duke Professor Cynthia Rudin is one of 20 women identified as being the most productive female researchers in AI, as measured by the number of papers posted on arXiv.org.
     
  • Location and research domain are significant drivers of gender diversity. Women in the Netherlands, Norway and Denmark are more likely to publish AI papers while those in Japan and Singapore are less likely to do so.  
     
  • There is a significant gender diversity gap in universities, big tech companies and other research institutions. Every academic institution and organization in the Nesta dataset - other than the University of Washington -  has less than 25% female AI researchers. In big tech, only 11.3% of Google’s employees who published AI research are women. The proportion is similar for Microsoft (11.95%) and slightly better for IBM (15.66%).