When we introduce artificial intelligence into evaluation processes, we’re not just changing tools: we’re redefining what matters, who matters, and how value is measured.
This is the starting point for my speech at the webinar “Evaluation and Artificial Intelligence as a Key to Equal Opportunities,” on March 30, organized by the Italian Evaluation Association.
I focused on a question I consider crucial: what happens when evaluation criteria encounter systems that learn from data already affected by inequalities?
What implicit assumptions are translated into algorithms?
What effects do we produce when we automate evaluation?
And above all: how can we use AI not only to measure, but to reorient policies in a more equitable direction?
We discussed evaluation as a political as well as a technical space, where concepts like Algorithms and Data Feminism become operational tools, not labels.
For me, it was an opportunity to reflect on how to make evaluation processes more informed, transparent, and transformative.
