AI is currently a straight white man and that is a big problem
Annie Romein-Verschoor Lecture image: AI-generated image
Artificial Intelligence (AI) is a technology that often puts women and minority groups at a disadvantage because it tends to be developed by straight white men. What if that changed and women were the driving force behind AI? This is the thought experiment at the heart of Maaike Harbers’ Annie Romein-Verschoor Lecture on 7 March.
Failings
The Dutch child benefits scandal, anti-cheating software that does not work properly for people with dark skin and an AI recruitment process at Amazon that rejected qualified women... Harbers, Professor of Artificial Intelligence & Society at Rotterdam University of Applied Sciences, reels off a list of examples of how the unchecked use of AI can lead to significant failings.
Unsurprising, says Harbers, because technology is a reflection of our unequal society. ‘Take the healthcare system where AI is increasingly used to make predictions about people’s health. This AI is fed with research data, and historically much more research has been conducted on men’s health, which means this technology does not work as well for women.’
Ethics and world views
So if AI does not use neutral data, AI itself cannot be neutral either. To compound this, AI is mostly created by a homogenous group. ‘Developers at big tech companies are generally straight, white and male. Their choices determine how the algorithm ends up working. They build their ethics and world view into code and consequently into the AI used by millions of people.’
Harbers gives the example of ‘a professor’. Ask ChatGPT to generate an image of a professor and this is almost guaranteed to be an image of a white man. ‘This perpetuates existing stereotypes in our society.’
‘My research shows, for example, that women are more committed to resolving ethical issues such as disinformation, privacy violations and data bias’ - Maaike Harbers
What if AI was female-dominated? Would AI be different? ‘Definitely!’ says Harbers. ‘My research shows, for example, that women are more committed to resolving ethical issues, such as disinformation, privacy violations and data bias. Women and minorities at tech companies who are committed to resolving ethical issues are often seen as party poopers, as putting the brakes on technology. When it’s hard enough as it is for them at these male-dominated companies.’
America and the power of big tech
Harbers is deeply concerned about the power of big tech, especially in America. Billions of people depend on the software these largely unregulated companies develop. And Trump’s election victory has fuelled an increase in alpha male behaviour within these companies. As Meta boss Mark Zuckerberg put it, ‘We need more masculine energy.’
‘This makes the tech environment even less attractive to women and minorities, leaving an even more homogenous group.’
International Women’s Day is essential
Harbers sees it as a great honour to be giving the Annie Romein-Verschoor Lecture, an annual lecture around International Women’s Day. ‘Women still do not have the same opportunities as men in this world and we are judged differently.’ She gives the example of someone commenting on her ‘lovely smile’ after she had given a presentation at a conference rather than on what she had to say. ‘Women are still the second sex in many respects.’
More diversity within AI
As well as a great honour, giving the lecture is also a great responsibility, says Harber, because she wants to show how much influence a homogenous group of developers has on AI and thus on society. ‘In my lecture, I ask what would happen if women had the upper hand in AI development. That is not a goal in itself, but more diversity in the world of AI is.’ This makes it important to show that the AI systems that currently dominate are only versions of AI.
‘It’s up to us to come up with ways to change AI and make it more ethical’, is her hopeful conclusion. ‘My students think it’s normal for responsible tech to be part of their AI programme. They don’t all find it equally interesting but they do understand it is important, which is a huge win.’
Annie Romein-Verschoor Lecture 2025
Maaike Harbers is giving her Annie Romein-Verschoor Lecture Betavrouwen en alfamannen: over de rol van gender in AI (STEM women and alpha males: on the role of gender in AI) on Friday 7 March. Sign up for the lecture (in Dutch).