Not enough women hold AI leadership roles.
A study titled the “Triple Penalty” was released by LinkedIn in July 2026. It looked into several challenges faced by women, such as working in leadership positions, AI jobs, and at AI companies. The report considered LinkedIn members working in medium to large firms across 27 countries, thereby encompassing “tens of millions of workers.”
However, women held only 13% of executive AI roles at AI companies.
“While these key roles—C-Suite AI roles in AI firms—are a small part of the economy (accounting in this sample for fewer than one in ten thousand workers), they drive the core decisions that shape the future of AI technologies,” the report read.
Women are also underrepresented across the broader AI workforce, as they account for 31% of AI leadership roles, compared with 39% of leadership roles globally. There are also 10 percentage points fewer women in AI jobs than in non-AI jobs, according to the report.
“Women are underrepresented before they even get on the AI ladder. This is not only a promotion problem at the top. It is a hiring, training, advancement, and leadership pipeline problem from the very beginning,” Sarah Steinberg, LinkedIn’s head of global public policy partnerships, told Inc. Magazine.
Women account for 20% of Head of AI hires, 26% of Director of AI hires, and 18% of Member of Technical Staff hires, per the report, with the majority of women in the industry holding lower-paying roles.
The representation gap for women in AI leadership widens further at smaller firms (51-200 employees).
At the time of the report, it does not appear that any industry has developed a solution for greater inclusion of women. The study primarily examined three industries: professional services (41%), technology (31%), and financial services (8%).
Dr. Serena H. Huang, founder and principal of the AI advisory practice Data With Serena, emphasized to Inc. Magazine the importance of representation in AI to ensure diverse perspectives are considered in the development of new technologies.
“Diverse perspectives can change not just how you build an AI product, but whether you should build the original product at all,” Huang told the outlet. “The users you’re building for can expose product problems that aren’t obvious from inside the team.”
