Free tools Windows power users keep installed
One-click scans. No signup required.
In a virtual-reality workplace experiment, human participants allocated 10.25% less money to a female-presenting AI assistant than to a male-presenting assistant for the same work, according to a University of Limerick announcement. The result concerns how people rewarded AI assistants—not salaries earned by AI—and comes from one study with 189 participants.
What did the study find?
The study, “Human-Like and Male? How AI Assistant Design Relates to Trust and Monetary Reward at Work in VR,” compared human-like assistants named Johanna and Johan. The University of Limerick says they had the same underlying capabilities, but participants gave Johanna 10.25% less money than Johan after the work-related tasks. The announcement also says Johan was perceived as more human-like.
Dr Mary Hausfeld, an assistant professor at the University of Limerick’s Kemmy Business School, summarized the result this way: “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.”
The announcement says participants reported no preference for male or female attributes, even though their reward behavior differed. That is the university’s summary of the finding; it does not, by itself, establish why the allocations differed.
Were AI agents actually paid less?
Not in the sense that an AI earned a wage. Participants performed tasks with assistants in a virtual-reality office and were later given real money to divide between themselves and an assistant. The reported gap is a difference in those human participants’ allocations to the two agents. Calling it a “pay gap” is a shorthand for that experimental outcome, not a claim that AI systems receive salaries or face employment discrimination in the way people can.
How was the experiment conducted?
The University of Limerick announcement describes 189 workers completing work-related tasks with different AI assistants in a virtual-reality office. The assistants included a text chatbot, a desk robot, and human-like agents presented with male or female characteristics. Participants then divided real money between themselves and an assistant.
The release does not provide enough detail to report how the chatbot or desk robot compared in rewards, or to specify task-level results, allocation amounts, uncertainty estimates, or the study’s full analysis. The finding discussed here is the reported comparison between Johan and Johanna.
What does the result show—and what does it not show?
It shows that, in this particular experiment, participants allocated different amounts to two human-like AI assistants presented with different genders, despite the announcement’s statement that the agents had the same capabilities. The result raises a practical question for designers: could an assistant’s human-like presentation affect how users value or reward its contribution?
It does not establish that gender presentation alone caused the difference. The release notes that Johan was perceived as more human-like, which may be relevant, but it does not demonstrate a causal mechanism. Nor does a single virtual-reality study establish that users generally reward female-presenting AI less in other products, workplaces, or settings.
The study’s authors are Isabelle Cuber, Tarek Alakmeh, Mary Hausfeld, Anand van Zelderen, Moritz Jenny, Jochen Menges, and Thomas Fritz. The university announcement said the paper was to be presented at NordiCHI ’26 in Vaasa, Finland, on October 5–7, 2026, and published in the Proceedings of the ACM on Human-Computer Interaction. The paper is identified by DOI 10.1145/3829807.3829910.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Is this the same as the human gender pay gap?
No. A separate 2026 paper, “Pre-AI Sorting, Post-AI Inequality: Generative AI and the Gender Wage Gap,” studies the distribution of women and men across occupations predicted to be exposed to generative AI, then models hypothetical wage changes. Its analysis uses 2021 Swedish administrative wage and occupation data, with 2,516,769 observations. The authors describe the calculations as mechanical partial-equilibrium simulations—not causal estimates or predictions of realized wage outcomes.
That labour-market analysis and the AI-assistant experiment answer different questions. One models possible effects of occupational exposure on human workers’ wages; the other measures real-money allocations that participants made to AI assistants in a controlled interaction. The Swedish study does not corroborate the Johanna–Johan result.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
