Have you heard of these individuals: Craig Dalton, Fei-Fei Li, Mira Murati, and Susan Millner? From the title of this piece, you might infer that they have something to do with the gender gap in AI—and you’d be, well, partially right.
For the past few years, the media has highlighted that fewer women than men use AI tools. But does the gender gap in AI still exist? Two new reports, released just two days apart last week, came to seemingly contradictory conclusions. Both drew on surveys of more than 5,000 U.S. adults conducted in 2026.
A Pew Research Center report noted that the gender gap in the use of AI chatbots, such as ChatGPT, has closed and that women’s use of AI has increased steeply over the past two years. According to the report, similar shares of women (47%) and men (50%) now use AI chatbots.
Another report, released two days earlier by the Shorenstein Center on Media, Politics and Public Policy at the Harvard Kennedy School, concluded that women (29%) were more likely than men (22%) to report that they had never used an AI tool.
How can we make sense of the discrepancy between these reports? The wording of the surveys—“AI chatbots” versus “AI tools”—may have made a difference. The surveys also used different sampling methodologies. The Pew Research Center, for example, used a nationally representative panel of U.S. adults.
But when I dug deeper into the reports, I found that their conclusions were more similar than different. The Shorenstein report identified no statistically significant gender difference in whether people had ever used AI when the researchers examined desktop browsing data, which measured people’s actual behavior.
But has the gender gap in AI really closed? Asking whether someone has ever used an AI tool may not be particularly meaningful. After all, a person could try an AI tool, decide they don’t trust or like it, and never use it again. A more useful metric may be whether women and men use AI regularly.
On this point, the reports agree. The Pew report found that men use AI chatbots daily more often than women do (27% versus 20%). Likewise, the Shorenstein report found that 25% of men used AI regularly, compared with only 16% of women. So, when it comes to regular or daily use, a gender gap persists. But why?
Here again, both reports provide useful insights. According to the Pew report, women are more likely than men to believe that AI will negatively affect them personally and society more broadly. Likewise, the Shorenstein report found that women trusted AI less than men did and were more likely to associate it with anxiety and worry. Overall, women expressed more negative attitudes toward AI than men did.
Gender Inequities Persist
The Shorenstein report provides additional evidence that helps explain the gender gap in AI use. Among employed respondents, women were more likely than men to report receiving no AI training at work (64% versus 49%). Women also reported receiving less mentoring and encouragement related to AI use.
Both women and men in the Shorenstein survey perceived men as better represented than women in AI news, development, and content. Men also dominated news coverage and public discussions about AI. Across 78,542 U.S. news articles, men appeared in headlines 14.6 times as often as women. Among experts cited in those articles, men were cited 4.3 times as often as women.
This brings us to the most stunning findings in the Shorenstein report. Participants were asked whether they recognized the names of several influential figures in AI development. The list included both women and men. For every name on the list, men were more likely than women to report recognizing the person.
Technology leaders Elon Musk and Mark Zuckerberg were recognized by more than 90% of respondents. By contrast, Daniela Amodei—the most widely recognized woman on the list—was recognized by only 49%, slightly less than Jensen Huang and Dario Amodei, the least widely recognized men, at 50% each.
Unbeknownst to the respondents, however, the researchers had slipped two fictitious names into the list of AI experts: Craig Dalton and Susan Millner. And here’s the kicker: Fei-Fei Li and Mira Murati, two influential women in AI development, were each recognized by only 46% of respondents—the same percentage who reported recognizing the two fabricated names. This finding underscores how invisible even some of the most influential women in AI remain to the public.
The Takeaway
The take-home message from these reports is not that women should be encouraged to use AI more often. There are likely good reasons to be cautious about using AI and its potential negative effects. The greater concern is the broader systemic gender inequity. We need to ask hard questions about whether women have equal opportunities to understand AI, shape its development, and influence the public conversation surrounding it.
The Shorenstein report explicitly recommends that organizations and employers do more to provide women with the support and information they need to decide how and when to use AI in the workplace. Journalists and news organizations can also widen the pool of experts they consult. If women do not see themselves represented in news about AI, is it any surprise that they use AI less regularly and remain more skeptical about its benefits?
Reflections on Fabricated Names
I’ll conclude with a broader point that extends beyond the gender gap in AI. Almost half of the Shorenstein survey respondents reported recognizing the fabricated names Craig Dalton and Susan Millner as AI experts. And men (53%) were more likely than women (40%) to make this claim, even though the two names are wholly fictitious.
Wow. How do we even make sense of that?
I suspect that in a culture that prizes certainty and confidence, admitting that one does not recognize an influential AI figure can feel like a weakness. And perhaps men struggle with this problem more than women do.
But acknowledging uncertainty and recognizing the limitations of one’s knowledge should be viewed as strengths. These are markers of intellectual humility—an attribute that research has shown helps people make sound decisions, resist misinformation, and relate better to others.
I don’t know. I could be wrong. I need more time to figure this out. These words may not project confidence. But they embody the qualities we need in a rapidly changing, AI-shaped world in which bold assertions often outpace genuine understanding.
If you appreciated this piece, you might also enjoy my newsletter on the science and practice of gratitude.