Men claim familiarity with fake AI executives, Harvard survey finds

A survey of more than 5,400 Americans by Harvard researchers found that 73% of men said they knew Sam Altman, OpenAI's chief executive. 53% also claimed to recognize Mira Murati, Susan Milner and Craig Dalton. Murati is real — former CTO at OpenAI and now chief executive of Thinking Machine Labs — but Milner and Dalton are entirely fictitious. 40% of women in the survey likewise claimed familiarity with the invented names, while 38% said they knew Murati.

Men were wrong at the same rate but were less likely to admit it. Laura Manley, director of the Shorenstein Center at Harvard Kennedy School, said men in the survey displayed higher optimism and confidence across every measure. The researchers found that men use AI tools more frequently and answer more factual questions correctly, yet the error rate for both genders was roughly the same. The gap came from women selecting "not sure" on average 3.3 of 12 questions, compared with 2.3 for men. It is the tendency toward artificial certainty, not superior knowledge, that inflates the results.

Men dominated media coverage and expert citations. In an analysis of a full year of AI news, men appeared in headlines 14 times more often than women. When the researchers isolated only people quoted as subject-matter experts, three-quarters were men — a share significantly higher than women's roughly 29% representation in the field's actual talent pool. Manley suggests it may be time to put more women forward as prominent experts, or at least make room for a voice that says "not sure."

Both genders ranked "concern" as the leading emotion toward AI, but women reported higher levels of apprehension, fear and worry. Men, by contrast, rated every positive emotion — curiosity, enthusiasm, excitement — higher. In the current discourse, Manley said, "there is so much confidence in either doom or abundance that it drowns out the voices that should be saying 'not sure.'" Many open questions remain about appropriate uses and the technology's future; one gender is simply less willing to acknowledge the uncertainty.