A relentless stream of movies, from Iron Man to Ex Machina, has helped entrench systemic gender inequality in the artificial intelligence industry by portraying AI researchers almost exclusively as men, a study has found.
The overwhelming predominance of men as leading AI researchers in movies has shaped public perceptions of the industry, the authors say, and risks contributing to a dramatic lack of women in the tech workforce.
Beyond the impact on gender balance, the study raises concerns about the knock-on effects of products that favour male users because they are developed by what the former Microsoft employee Margaret Mitchell called “a sea of dudes”.
“Given that male engineers have repeatedly been shown to engineer products that are most suitable for and adapted to male users, employing more women is essential for addressing the encoding of bias and pejorative stereotypes into AI technologies,” the report’s authors write.
Researchers at the University of Cambridge reviewed more than 1,400 films released between 1920 and 2020 and whittled them down to the 142 most influential movies featuring artificial intelligence. Their analysis identified 116 AI professionals. Only nine of these were women, of which five worked for a man or were the child or partner of a more senior male AI engineer.
The study highlights the Avengers film franchise , which depicts a stereotypical lone male genius (Tony Stark, aka Iron Man) who has mastered so many skills that he can synthesise an element and solve the problem of time travel “in one night”. In Alex Garland’s 2014 movie Ex Machina, another lone genius is so successful that he rises above the norms of ethics and law to subject an employee to violence while amusing himself with sex bots.
The earliest film in the list with a female AI creator is the 1997 movie Austin Powers: International Man of Mystery, in which a shouty Frau Farbissina unveils a trio of “fembots” fitted out with bullet-firing breasts.
Dr Kanta Dihal, a co-author on the study and a senior research fellow at the Leverhulme Centre for the Future of Intelligence, said part of the male bias was an “art-mimicking life” spiral whereby film-makers portray AI professionals as men to reflect the male dominance of the industry. But about one in five AI engineers are women, compared with less than one in 10 of those portrayed in cinema. “They are exacerbating the stereotype they see,” she said.
The lack of female AI engineers on screen may also be linked to the dearth of women behind the camera. According to the study in Public Understanding of Science, not one prominent film about AI in the past century was directed solely by a woman. The study is accompanied by a report posted on the researchers’ website.
Dihal believes the perpetuation of male stereotypes is damaging on several levels. First is the impact on career choice, with women potentially dissuaded by the perception that AI is only for men. Second is the effect on hiring panels, who might come to perceive men as a better “cultural fit” for a tech firm. Then there is the office culture. “If a female AI researcher gets into the workplace, what kind of stereotypes and assumptions is she going to have to contend with?” Dihal said.
Prof Dame Wendy Hall, a regius professor of computer science at the University of Southampton, said there was an urgent need for a campaign to increase diversity in AI. Hall wrote her first paper on the lack of women in computing in 1987, and said the situation was worse with AI because the potential impact on society was so great.
“Clearly the media hugely influences the decisions young people make about their future careers,” she said. “If they perceive AI as a male-dominated profession, this will make any other attempts to rectify the current situation that much harder. The problem of course is that there are no quick fixes, as the many attempts to attract more women into computing have sadly shown us.”
Prof Judy Wajcman, an emeritus professor of sociology at the London School of Economics and principal investigator on the Women in Data Science and AI project at the Alan Turing Institute, said: “Male-dominated images in popular culture are bound to deter women from entering the field. Key here is the way in which hi-tech leaders are represented as genius visionaries, reinforcing the idea that women are not cut out for the field.”
She added: “I strongly endorse the authors’ call for a substantial increase in the cinematic portrayal of women in AI. But equally we need to change the reality the films reflect. That is, to increase diversity in AI leadership roles, and especially the ‘tech bro’ culture which makes it difficult for women to flourish in this sector.”
• This article was amended on 13 February 2023 to clarify that the study highlighted the skills mastered by the Iron Man character in the Avengers film franchise generally not just in the 2008 movie Iron Man. The headline was amended on 14 February 2023 to clarify that the numbers apply to key films, not all films.