Gender bias in the workplace can be replicated in interactions with artificial intelligence, new research from the University of Limerick (UL) has found.
The study examined whether the way AI assistants are presented influences how people treat them in a workplace setting.
It was co-authored by Dr Mary Hausfeld of UL’s Kemmy Business School.
Researchers placed 189 workers in a virtual reality office and asked them to complete work-related tasks with different AI assistants, including human-like agents presented with either male or female appearance.
READ MORE
Participants were then given real money to divide between themselves and the AI assistant for completing the tasks, allowing the researchers to “recreate a real-world transaction and examine how much people were willing to pay AI for its work”.
AI agents with a female appearance were paid less than their male-presenting counterparts, despite using the same underlying technology, according to the research.
The AI agents had the same underlying capabilities, yet participants treated them differently depending on how they were presented.
The male-presenting AI agent, Johan, received higher monetary rewards than the female-presenting agent, Johanna. The male agent was also perceived as more human-like.
Johanna was paid 10.25 per cent less than Johan for completing the same work.
The researchers concluded that their findings suggest gender biases familiar from human workplaces could be reproduced in interactions with AI.
Hausfeld said the results raise important questions as AI agents increasingly become part of people’s working lives.
“What is striking about our findings is that the technology behind these AI agents was exactly the same, but people did not treat them in the same way.
“We often think of AI as being neutral, but the way we design and present these systems can activate those same assumptions and biases that exist in our interactions with other people.
“As AI agents become more common in the workplace, we need to think carefully about the characteristics we give them and the behaviours those choices may encourage. We don’t want to inadvertently reproduce existing inequalities in a new technological setting.”
The study was part of a larger investigation into how the human-like appearance of AI influences workplace interactions.
Although the systems were powered by the same underlying AI, participants trusted the human-like assistants more and gave them more credit for their contribution than the robot.
Participants also suggested they had no preference for an AI with female or male attributes, but “their behaviour told a different story”.
The study was conducted by Isabelle Cuber and Tarek Alakmeh of the University of Zurich, Mary Hausfeld of University of Limerick, Anand van Zelderen of SKEMA Business School, and Moritz Jenny, Jochen Menges and Thomas Fritz of the University of Zurich.
Human-Like and Male? How AI Assistant Design Relates to Trust and Monetary Reward at Work in VR will be presented at the 14th Nordic Conference on Human-Computer Interaction in Vaasa, Finland, from October 5th to 7th.
It will also be published in the conference proceedings by the Association for Computing Machinery.
