The study, conducted by Harvard researchers, interviewed more than 5,400 American respondents. Among men, 73 percent said they recognized OpenAI chief Sam Altman, while just over half,53 percent,claimed to know Mira Murati, Susan Milner and Craig Dalton. Murati is a real executive, currently leading Thinking Machine Labs, but Milner and Dalton have no existence beyond the questionnaire, highlighting a gap between confidence and factual knowledge.
Men reported using AI tools more often than women and, on average, answered factual AI questions more correctly. However, the overall error rate was similar across genders. The study measured uncertainty by counting how many of twelve questions respondents marked as “not sure.” Women indicated uncertainty on an average of 3.3 items, compared with 2.3 for men, suggesting a higher propensity to express doubt among female participants.
Analysis of a year’s AI news headlines revealed that male names appeared more than fourteen times as often as female names. When the researchers narrowed the focus to quoted experts, three-quarters of the sources were men, a share that exceeds the estimated 29 percent representation of women in the AI talent pool. This imbalance indicates that media outlets amplify male voices beyond their proportion in the field, potentially shaping public perception of expertise.
Both sexes identified anxiety as their primary emotional reaction to AI, yet women reported higher levels of concern, fear and worry. Men, in contrast, assigned greater intensity to positive emotions such as curiosity, enthusiasm and excitement. The divergent affective profiles suggest that while the technology provokes unease across the board, gender influences the balance between apprehension and optimism.
Researcher Manley argues that the platform should feature more women who are willing to voice uncertainty. “Maybe there should be more women prominently featured as experts or given the platform to express: not sure,” Manley says, as reported by WIRED. The comment underscores a call for balanced representation that includes candid acknowledgment of unknowns rather than a chorus of unqualified certainty.
The findings reveal a gendered confidence gap that permeates both personal knowledge and public discourse on AI. As unanswered questions about appropriate applications and future impacts multiply, the tendency of one gender to suppress doubt may limit critical debate. Addressing this imbalance could foster a more nuanced conversation, encouraging diverse perspectives to surface and reducing the dominance of overconfident narratives in shaping AI policy and development.