Progress by design: Tackling gender bias in AI.




Artificial Intelligence (AI) is a massive opportunity for humanity, reshaping how we work, learn, and live at gargantuan speed. But AI doesn’t operate in a vacuum, it “learns” from the world we have built. If AI tools and systems are trained with biased data, stereotypes and views shaped by decades of inequality, these will influence the final product.

Currently, only 28 per cent of people are aware of gender bias in AI. Once informed, over half said they were concerned, especially Gen Z and Millennials. And they’re right to be. AI systems have already failed repeatedly: facial recognition software has led to false arrests, AI hiring systems have favoured male applicants over women, and healthcare algorithms have misjudge clinical needs—especially for patients from minority groups—leading to missed care that could save lives.

The problem isn’t just who uses AI; it’s who builds it. Women currently make up just 35 per cent of employees in US tech firms yet only 40 per cent of people surveyed by Team Lewis see the lack of women in AI leadership as a problem. And just 24 per cent believe AI frequently portrays women in misogynistic ways. This points to a deeper issue—when AI learns from a world shaped by inequality, it risks encoding that inequality into the future.

Still, there is growing demand for change and 66 per cent of people believe governments should regulate AI to reduce gender bias. That means tackling everything, from image generators that produce stereotypical visuals, to algorithms that penalize CVs with female indicators. Many are also calling on companies to invest in inclusive design teams, ethical standards, and better training data.

Comments

Popular posts from this blog

Analyzing the Gender pay gap in the United States.

Progress by design: Tackling gender bias in AI.

10 Años de Lucha - Honoring 10 Years of the Latina Equal Pay Campaign | PVBLIC Foundation.