Cybersecurity for Artificial Intelligence & Machine Learning


Units: 6


Advancements in Artificial Intelligence (AI) and Machine Learning (ML) have allowed for a surge in adoption of AI & ML solutions to address problems across numerous domains. With this rising reliance on AI & ML in many organizations, it is critical that such systems are protected from malicious activities. This course will discuss AI & ML cybersecurity issues, explore case studies of AI & ML cyber incidents, present AI & ML adversarial techniques, and demonstrate secure design approaches to protect AI & ML systems. With an emphasis on machine learning, the course will focus on secure machine learning systems development approaches and secure machine learning operations (MLOps). Students are expected to have knowledge of fundamental statistics and the ability to program in Python.

Learning Outcomes

  • Cybersecurity Foundations: Confidentiality/Integrity/Availability; Cyber Kill Chain; Threat Actors & Threat Modeling; Cybersecurity Frameworks 
  • ML Fundamentals: Supervised, semi-supervised, Unsupervised and Reinforcement Learning; Neural Networks & Deep Learning; Regression, Classification, Clustering, and Anomaly Detection 
  • Adversarial Machine Learning:
  • Data Security: Data Curation; Data Poisoning; Label Flipping; Input Manipulation Attacks, 
  • Model Security: Model Extraction; Membership Inference Attacks; Model Inversion Attacks; Model Supply Chain Attacks;
  • AI/ML Defenses & Mitigations: Data Sanitization; Input Sanitization; Model Inspection; Data Encryption; Adversarial Training
  • Generative AI Security: Protecting LLMs; LLMs for Secure Code Generation; LLM Cyber Risks; Deepfake Media; Digital Authenticity
  • Secure MLOps: MLOps; AI/ML Software Security; Data Versioning; Model Versioning; Model Deployment; Model Operation & Model Monitoring

Prerequisites Description

90-812 Python Programming I <or>

 95-888 Data Focused Python <or>

 95-898 Introduction to Python 



 90-707 Statistical Reasoning <or>

  90-711 Statistical Reasoning with R