HDIW | SAFE AI - Paolo Giudici

Paolo Giudici, a Professor of Machine Learning and Artificial Intelligence at the Florence School of Banking and Finance and the University of Pavia, discusses measuring the opportunities and risks of AI using AI itself. He defines AI as machine-based systems that autonomously convert inputs into outputs, influencing both physical and digital environments. While AI brings novel opportunities, it also introduces significant risks, prompting the development of global regulations.

Paolo introduces a model inspired by the European AI Act, structured around four key principles that form the acronym SAFE:

- Sustainability
- Accuracy
- Fairness
- Explainability

This model allows AI to measure its own risk levels, creating a comprehensive risk assessment. The resulting score helps identify how risky an AI application is: the higher the score, the greater the risk.

---

Follow EU-SDFA activities: