Men and women in the United States are now about equally likely to say they have used an artificial intelligence chatbot, but women are more likely to worry about the technology's pace and future effects, according to a new Pew Research Center breakdown of an earlier national survey.
Pew's Oct. 16 analysis draws on interviews with 5,119 U.S. adults conducted Feb. 17-23. It is a closer look at data from a broader study released in June, not a new October survey. The findings do not include a separate New York or Finger Lakes estimate.
Use converges, frequency differs
Forty-nine percent of adults said they had ever used an AI chatbot such as ChatGPT, Gemini or Copilot. The shares were 50% of men and 47% of women, a much smaller gap than Pew found in 2024, when 39% of men and 28% of women reported use. Pew changed the question's wording and who was asked between the two surveys, a qualification that limits a simple trend comparison.
Daily use still differed: 27% of men and 20% of women said they used chatbots every day. ChatGPT use was identical among men and women, at 44% each. Men were more likely to report ever using Copilot, 22% to 13%, and Gemini, 29% to 20%.
The survey asked respondents about a range of tasks. Among all adults, 45% of men and 39% of women said they had used a chatbot to search for information. For work tasks, the shares were 40% and 35% among adults employed for pay. Those figures should not be treated as percentages of chatbot users or of all women and men, respectively, in the work question.
Women were slightly more likely to report using a chatbot for emotional support or advice, 11% to 8%. Both sexes reported the same 4% rate for companionship. Pew said men and women generally cited similar uses, so the differences should not be read as wholly separate patterns of behavior.
Thirty-five percent of men said chatbots helped their productivity a lot or a little, compared with 25% of women. Pew found smaller or no gender differences in perceived effects on knowledge, creativity, happiness or personal relationships. The responses measure people's assessments, not independently tested productivity gains.
Women more concerned about impact
Asked about AI's effect on them personally over the next 20 years, 33% of women expected a negative impact and 17% a positive one. Men were nearly evenly divided, at 27% negative and 29% positive. The remaining respondents anticipated mixed effects, were unsure or did not answer.
On society as a whole, 43% of women expected a negative impact and 11% a positive one. Men also leaned negative, 36% to 22%, but by a smaller margin. These are predictions about the future, not measurements of AI's effects to date.
Sixty-eight percent of women and 58% of men said AI was advancing too quickly. Men were more likely to say it was moving at about the right pace, 23% to 14%. Very few in either group said it was moving too slowly.
Fifty-five percent of men and 41% of women said they had heard a lot about AI. Men were also more likely to report being extremely or very confident using chatbots, 22% to 15%. Pew did not conclude that familiarity alone caused the difference in views.
Pew used its probability-based American Trends Panel and weighted responses to represent U.S. adults. The survey methodology describes how panelists were recruited and interviewed. As with other surveys, the reported gender patterns reflect respondent answers at one point in time, not causal proof about why any individual uses or distrusts AI.






