Visualizing AI biases through language errors

In this workshop, we will explore how spelling mistakes, translations, and other linguistic errors can reveal hidden social biases within generative AI systems. We will begin with a bit of theory to understand how these biases are formed, followed by a practical session where we will use open-source tools to identify them. Finally, we’ll reflect on how language can mirror societal inequalities and the role technology plays in this dynamic. This is a unique opportunity to uncover how language can serve as a lens to see how AI perpetuates stereotypes and social inequalities.

Organized by: Digicoría, collaborative project.