SPAIML 2025
Security and Privacy-Preserving AI/ML 2025

Table of Contents

• Preface
  Summary: There were 10 papers submitted for peer-review to this workshop. Out of these, 7 papers were accepted for this volume, all of which as regular papers.

Session 1: Paper Presentations (1)

• Alignment and Adversarial Robustness: Are More Human-Like Models More Secure? 1-10
  Blaine Hoak, Kunyang Li, Patrick McDaniel
• Collaborative Reinforcement Learning for Cyber Defense: Analysis of Strategies, and Policies 11-20
  Davide Rigoni, Rafael Fernandes Cunha, Frank Fransen, Puck de Haan, Amir Javadpour, Fatih Turkmen
• Experimental Evaluation of Non-Natural Language Prompt Injection Attacks on LLMs 21-30
  Huynh Phuong Thanh Nguyen, Shivang Kumar, Katsutoshi Yotsuyanagi, Razvan Beuran
• Preliminary Investigation into Uncertainty-Aware Attack Stage Classification 31-41
  Alessandro Gaudenzi, Lorenzo Nodari, Lance Kaplan, Alessandra Russo, Murat Sensoy, Federico Cerutti

Session 2: Invited Talk

• Abstract: Should I trust it with my data? Capabilities, limits, and perspectives of AI technologies for privacy 42
  Victor Morel

Session 3: Paper Presentations (2)

• Language Models in Cybersecurity: A Comparative Approach to Task-Driven Model Assessment 43-54
  Holger Schmidt, Klaus Kaiser, Daniel Spiekermann
• Seed Scheduling in Fuzz Testing as a Markov Decision Process 55-67
  Rafael Fernandes Cunha, Luca Müller, Thomas Rooijakkers, Puck de Haan, Fatih Turkmen
• TAIBOM: Bringing Trustworthiness to AI-Enabled Systems 68-78
  Vadim Safronov, Anthony McCaigue, Nick Allott, Andrew Martin