In this powerful episode, we explore Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification, a groundbreaking study authored by Joy Buolamwini and Timnit Gebru. This work shines a spotlight on how AI systems, widely used in facial recognition, disproportionately misclassify women of color. Using an intersectional framework, Buolamwini and Gebru reveal the deep flaws in commercial AI systems, with error rates of up to 34.7% for darker-skinned women, while lighter-skinned men show error rates as low as 0.8%. Their research exposes the urgent need for more inclusive and ethical AI design, and it has sparked a global conversation about bias in technology.

Join us as we discuss how this research is pushing tech companies to rethink algorithmic fairness and accountability, and the steps we can take to build a more equitable future in AI.

Credit: Joy Buolamwini and Timnit Gebru
(Source: Proceedings of Machine Learning Research 81, 2018)

Tune in to learn how this work is revolutionizing the way we approach ethics in AI, and why it’s a message the world needs to hear.