A new Belgian study says gender bias in AI-assisted recruitment is often underestimated, according to The Brussels Times.
The research was done by the Institute for the Equality of Women and Men, together with the University of Liege and Hasselt University. It looked at how organisations in Belgium use artificial intelligence when hiring.
The study found that about three out of four recruiters now use AI at some stage of the hiring process, such as screening applications or reviewing CVs.
As per the news report, many recruiters think that AI is fairer than human decision-making. But the study shows that AI can copy existing stereotypes and create discrimination, making inequalities worse rather than better. Only a small number of recruiters - roughly 12-17% - said they had seen biased outcomes from AI systems. Most of the bias they noticed was linked to social media algorithms, age and ethnicity. Very few mentioned gender bias on their own.
Also Read | Your Moltbook Questions, Answered: What The Platform Is, And What It's Not
The Institute warned that Belgium needs stronger legal protections for AI tools. It found that only one in five organisations have policies that match current AI laws. The European Union's AI Act labels recruitment tools as "high risk" and says they must be carefully checked and protected against bias.
Also Read | Real Or AI? The Story Behind 'Conjoined Twins' That Have Massive Social Media Following
The Institute also said Belgium should look at its anti-discrimination laws to make sure they work well for AI used in hiring and at work. Later this year, it will publish a brochure to help employers understand how to use AI responsibly. The study questioned over 400 recruitment professionals.
