Women trust and use artificial intelligence less than men, astudy released Oct. 14 by the Shorenstein Center on Media, Politics and Public Policy at the Harvard Kennedy School found.
Fifty-seven percent of women reported never having used AI or using AI tools rarely, compared with 43% of men in the study, which was based on a Civic Health and Institutions Project survey of U.S. adults in July.
The survey was backed up by anonymized browsing data from U.S. adults collected between July 2024 and August 2026 to match self-reported results with actual AI use.
“Women, pretty much across the board — from the workforce, encouragement, and training, to how they're represented in the media, to their trust — are consistently underrepresented,” Shorenstein
Center executive director Laura Manley said. “I was surprised that it was so, frankly, holistic across so many critical elements of the pipeline.”
An average of 6.4 out of 10 women reported feeling more concern associated with AI, compared with an average of 5.9 out of 10 men. Inversely, men reported feeling more curious about AI with an average rating of 5.8 out of 10 compared with 4.9 for women.
“As AI continues to expand, I see the trust getting worse, unless companies start hiring women into AI decision-making and design roles — and taking their input,” gender equity researcher Amy Diehl told USA TODAY.
Previous research has also found that women are less represented in AI and that it is more likely to be biased against women. For example, chatbots often rank males as more intelligent and more often show male leaders as role-models. Last year, women made up only 26% of AI hires in the United States and held just 13% of C-suite AI roles globally.
And on Oct. 13, the largest tech conference for women — the Grace Hopper Celebration — was canceled due to lack of sponsors and funding.
“The purpose of this report isn't to advocate for AI across the board and say women and everyone should be using it more,” Manley said. “But to surface the discrepancies in the underrepresentation and opportunity.”
Lack of media representation
The study also looked at over 75,000 U.S. news articles about AI between July 2025 and June 2026 to study how women are represented in the industry, finding significantly less representation for women.
“The more women are underrepresented in the media, the more difficult it is for women to see themselves represented and have an advocate on the other side,” Manley said.
Anthropic president Daniela Amodei was recognized 49% of the time, the most for any woman. The least recognized men — NVIDIA CEO Jensen Huang and Anthropic CEO Dario Amodei — were both recognized 50% of the time.
“I think if you blew this out and you looked at the tech industry more broadly, the numbers would be even more extreme,” Manley added.
Amazon CEO Jeff Bezos saw the most name recognition at 85%, while Stanford computer science professor Fei Fei Li, who is often cited as the “Godmother of AI,” was recognized 46% of the time.
In 2016, Li told Bloomberg Professional Services that “if you were a computer and read all the AI articles and extracted out the names that are quoted, I guarantee you that women rarely show up.”
Her comment holds true a decade later, according to the study. Men account for 93.6% of all headline mentions against 6.4% for women, and women made up fewer than 10% of sources in articles cited as authorities on AI, it found.
Women’s occupations are more exposed to automation by AI globally, at 4.7% compared with 2.4% of men’s. A report from the National Girls Collaborative Projects this year found that women only earn 23% of all computer science bachelor’s degrees.
“Women trust AI less than men and for good reason,” Diehl said. “Since they are largely not at the table when designing AI tools, they suffer disproportionate risks,” she said, citing privacy risks, occupational concerns and price increases.
“If we’re just using [artificial general intelligence] as a litmus test here, the bigger picture that society needs to take into account is about women and how they're playing a role in the development, the deployment, and the representation of emerging technologies,” Manley said.
Women more likely to be ‘unsure’
Use of AI tended to correspond to confidence in knowledge about the technology. Men were more likely to answer factual questions about AI correctly, averaging 7.1 correct answers out of 12 questions compared with women’s 6.3.
Rather than answering incorrectly though, women were more likely to respond “not sure,” at an average of 4 times out of 12 compared with 2.9 times out of 12 among men
“Uncertainty in selecting ‘not sure’ does not mean that women don't have the skills or they're not competent in AI,” Manley said. “Oftentimes when you hear about women being more skeptical, it gets conflated with their lack of ability somehow, or their internal perceptions of whether or not they would be good at using AI or something like that. They're just more skeptical, or they're less likely to say, ‘I have a hot take on this AI topic’ in the mainstream discourse.”
Men and women were each more likely to say their own gender was treated unfairly by AI, according to the study, with 20% of women saying women are more likely to be treated unfairly by AI and 17% of men saying men are more likely to be treated unfairly by AI.
Again though, women were more likely to be unsure whether AI treats any group unfairly, at 42% compared with 31% of men. Men were more likely to say that no group is treated unfairly.
“If you read this quickly, you could say, okay, well, it's a bad thing that more women aren't using AI. But there's lots of situations where AI isn't the right call,” Manley said, referencing recent criticism of AI use in the classroom as destructive for students.
Greta Reich covers the artificial intelligence industry for USA TODAY through a fellowship from the Tarbell Center for AI Journalism. Funders do not provide editorial input.
This article originally appeared on USA TODAY: Does AI really divide men and women? This study finds out
