Shocking: AI Is Learning Sexism and Racism From HumansHot Buzz

June 24, 2026 12:55
Shocking: AI Is Learning Sexism and Racism From Humans

(Image source from: Canva.com)

We often hear that AI is becoming more intelligent than humans, with people like Elon Musk claiming it already has some advantages. However, a recent warning from the UN indicates that AI may have picked up some of humanity’s worst prejudices. In a study of 133 AI systems, researchers found that about 44% displayed gender bias, and over a quarter exhibited both gender and racial biases. “Large language models (LLMs) often associate women with home life, family, and childcare, while relating men to business, leadership, and professional achievements. ” The UN also mentioned, “In some instances, AI generated responses that depicted women as sexual objects or inferior to men.” According to UN Women, when large language models were asked to finish sentences that started with a person's gender, around one in five replies were sexist or misogynistic.

Some responses even referred to women as possessions or objects, based on the research findings. The UN emphasized that as AI is increasingly used for tasks like writing emails, creating presentations, producing content, and answering questions for countless individuals, these inequalities are being reinforced by biased algorithms.

As per UN Women, this is not an error – it occurs because AI learns from internet sources filled with long-standing stereotypes, prejudices, and discrimination. Therefore, when AI studies human history, it may replicate human biases. Jayathma Wickramanayake, who leads Digital Technologies at UN Women, stated that AI models “extract biases from years of text created by people, about people, in a context where women were categorized under home and family, while men were classified under business and careers. ” The report pointed out that women already endure a heightened level of abuse online, and AI is making some types of violence easier to create and disseminate.

According to data from UN Women, nearly 25% of women who are human rights defenders, activists, and journalists reported facing violence online that was aided by AI. Furthermore, 12% noted that personal images were shared without their consent, while 6% indicated that they were targeted by deepfakes or altered images and videos.

UN Women advocates for gender equality, along with the rights and experiences of women and girls, to be included at every step of the AI development process, from creation to use and governance. Researchers also discovered racial bias in numerous systems, raising alarms that AI might perpetuate stereotypes affecting various groups, not just women. While we frequently discuss AI taking over human intellect, perhaps the more significant challenge lies in ensuring it doesn’t adopt humanity's worst habits and biases.

If you enjoyed this Post, Sign up for Newsletter

(And get daily dose of political, entertainment news straight to your inbox)

Rate This Article
(0 votes)