Algorithmic bias reproduces gender discrimination in digital era

Algorithms used in recruitment, advertising, and performance measurement are increasingly criticized for reproducing historical social inequalities. Because these systems learn from human-generated data, they often inherit existing class and gender disparities, treating past social imbalances as objective standards for future merit.

In recruitment, algorithms may favor resumes that mirror historically male-dominated leadership profiles, effectively automating the marginalization of women. Similarly, targeted advertising often reinforces stereotypes by directing ads for beauty and domestic care toward women, while presenting technology and investment opportunities to men. These patterns are driven by profit-seeking models that prioritize low-cost, high-profit audiences.

A June 2026 study by UN Women highlighted the scale of this issue, examining 133 artificial intelligence systems. The findings revealed that 44% of the systems studied exhibited gender bias, while 26% demonstrated a combination of both gender and racial bias.