University of Amsterdam

We are hiring: For the project "Ideology, Emotion Detection AI, & the Propagation of Social Inequality" we are looking for a post-doc (application deadline April 15th). The project examines how AI emotion detection models may perpetuate political ideology by reinforcing gender and ethnic stereotypes. A key concern is that these models are trained on datasets labeled by human annotators, whose political ideology may shape how they categorize emotional expressions—often in ways that align with stereotypes. When AI systems learn from these biased labels, their outputs can further influence human decision-making, unintentionally reinforcing existing inequalities.

To investigate these dynamics, the project will hire a post-doc for 12 months, starting this spring, see the vacancy: https://werkenbij.uva.nl/en/vacancies/postdoc-investigating-human-sources-of-bias-in-ai-face-classification-models-netherlands-13907