Skip to content

Why do algorithms replicate inequalities in gender and race?

We are at a turning point in history. The number of automated decision-making systems is unprecedented, and they are being deployed at high speed, but they do not guarantee equality at all. Español Português

Why do algorithms replicate inequalities in gender and race?
Google employees at its European headquarters in Dublin, Ireland, join others from all over the world who leave their offices in protest of allegations of sexual harassment, gender inequality and systemic racism in the tech giant. - Image by: Niall Carson / PA Archive / PA Images. All rights reserved
Published:

Today, the intersection of gender and technology is reproducing and even amplifying the patriarchal model. For example, of all the professionals who work on artificial intelligence (AI) only 22% are women. Google and Facebook reported that women working on AI represent the 10% and 15% respectively. In Latin America only 38% of internet users are women and, in the UK, women who work on cybersecurity earn 16% less compared to men.

If we consider race, the scene is even worse. In 2018 Google reported that 25,7% of women work in technical positions, but this number decreases to 0,8% if we consider black women. We see the same tendency in other technological giants, such as Facebook, Apple and Microsoft. This makes the diversity crisis within the technological and computer science ecosystems evident.

¿Why is this so relevant? Today there is a fast-growing increase in systems and apps that automate processes, using AI. Every day there are more countries implementing IA national strategies to promote its use, since it helps solve highly complex problems, makes automatization easier and makes a more efficient use of resources.