We have an opportunity to build systems that don’t just replicate our current inequities. Will we take them?
The notion that artificial intelligence could help reduce racism might seem counterintuitive. After all, we’ve seen numerous headlines about AI systems perpetuating or even amplifying racial biases. Yet as we enter 2025, amid both a backlash against social justice initiatives and the rapid proliferation of AI technologies, an unexpected opportunity is emerging.
Here’s the paradox: while few people want to be labeled racist, study after study reveal persistent racial disparities in everything from healthcare outcomes to economic opportunities. The fascinating twist? Society appears far more willing to critically examine and address bias in AI systems than confront human bias directly.
Sanmi Koyejo is an assistant professor of computer science at Stanford University. He also leads the Stanford Trustworthy AI Research (Stair).
About Author
More Stories
Scientists find ‘mutant’ gene behind foul-smelling species of wild ginger
The Guardian view on drought warnings: risks to the food supply need confronting | Editorial
Meta blocks major Muslim Instagram page in India amid rising conflict