Artificial Intelligence for Cybersecurity

Bojan Kolosnjaji is a researcher working at the intersection of artificial intelligence (AI) and cybersecurity. He has obtained his master's and PhD degrees in computer science from the Technical University of Munich (TUM), where he conducted research in anomaly detection methods in constrained environments. Bojan's academic work deals with anomaly detection problems in multiple cybersecurity-relevant scenarios, and the design of AI-based solutions to these problems. Bojan is currently working as a principal engineer in cybersecurity sciences and analytics, helping various cybersecurity teams deal with large-scale data, adopt AI practices and solutions, and understand security challenges in AI systems. Xiao Huang holds a doctorate in computer science from TUM. He is also a visiting scholar at Stanford University. His main research interests include adversarial machine learning (ML), reinforcement learning, anomaly detection, trusted AI, and AI applications in cybersecurity. Huang has published several top-tier conference and journal papers with over a thousand citations in both the ML and security domains. He led the ML research group at Fraunhofer AISEC Institute in Munich and also worked as a research scientist at Bosch Center for AI. He managed a data scientist team that designed and developed ML systems to tackle different cybersecurity problems. Peng Xu has focused on AI for system security, large language model (LLM) security, graph neural networks, program analysis, compiler design, optimization, and cybersecurity. He completed his master's at the Chinese Academy of Science in 2013 and pursued a PhD in IT security at TUM from 2015 to 2019. He is currently awaiting his dissertation defense. Peng's research topics include malware detection, private computation, and software vulnerability mitigation using compiler-based approaches. Peng is currently working as a principal engineer in compiler optimization and programming LLMs, especially on the topics of using LLMs to generate code blocks to detect malicious code as well as bug localization. Apostolis Zarras is a cybersecurity researcher with a rich academic background. He has served as a faculty member at both Delft University of Technology and Maastricht University. Dr. Zarras earned his PhD in IT security from Ruhr-University Bochum, where he honed his expertise in systems, networks, and web security. His research is driven by a passion for developing innovative security paradigms, architectures, and software that fortify ICT and IoT systems. Beyond his technical contributions, Dr. Zarras delves into the dark web and its underground markets, uncovering and combating malicious activities to bolster global cybersecurity. His work is dedicated to advancing IT security and protecting users and systems from emerging cyber threats.

Table of Contents

Big Data in Cybersecurity
Automation in Cybersecurity
Cybersecurity Data Analytics
AI, Machine Learning, and Statistics - A Taxonomy
AI Problems and Methods
Workflow, Tools, and Libraries in AI Projects
Malware and Network Intrusion Detection and Analysis
User and Entity Behavior Analysis
Fraud, Spam, and Phishing Detection
User Authentication and Access Control
Threat Intelligence
Anomaly Detection in Industrial Control Systems
Large Language Models and Cybersecurity
Data Quality and Its Usage in the AI and LLM Era
Correlation, Causation, Bias, and Variance
Evaluation, Monitoring, and Feedback Loop
Learning in a Changing and Adversarial Environment
Privacy, Accountability, Explainability, and Trust - Responsible AI
Summary