Hi! I am an Assistant Professor (RTD-A) at the Scuola Superiore Sant’Anna in Pisa, working on safe and secure artificial intelligence. I received my Ph.D. in Emerging Digital Technologies (2024), focusing on safety and robustness of AI for cyber-physical systems and computer vision, under the supervision of Prof. Giorgio Buttazzo and Prof. Alessandro Biondi. In 2023, I visited the Computer Vision Laboratory (CVLab) at EPFL with Dr. Mathieu Salzmann, studying robustness of self-attention and adaptive learning for autonomous driving.

I care about building AI we can trust, merging theory and practice to make models robust, transparent, and efficient in real deployments, especially when decisions affect safety. My current research spans adversarial robustness (stress-testing vision and NLP systems against evasion and poisoning attacks), efficiency under constraints (balancing accuracy, computation, and energy in distributed AI), explainability and uncertainty (interpreting model behavior and extracting reliable signals about failures), and out-of-distribution and risk-aware metrics (quantifying failure severity in safety-critical tasks such as autonomous driving).

Latest News!

πŸ† Oct 2025 β€” Recipient of the 2025 National Prize for Young Italian Researcher in Cybersecurity.
πŸ† Oct 2025 β€” Paper accepted in Internet Of Things on the robustness of distributed DNNs. Link
πŸ† Sep 2025 β€” Paper accepted in Pattern Recognition on the robustness of vision models. Link
πŸ“ Sep 2025 β€” New preprint on LLM Alignment with GTPO. See preprint
πŸ† Sep 2025 β€” Awarded the PhD Dissertation Prize from the Italian Society of Intelligence. Link

Contact

giulio.rossolini@santannapisa.it

Last update: September, 2025