300

This lab-intensive course examines the statistical, data-based nature of AI vulnerabilities. Students will simulate evasion attacks where input data is subtly tweaked to fool a model into an incorrect prediction and data poisoning where manipulated data is inserted into training sets.

Focuses on building security into the ML lifecycle through AISecOps and MLSecOps. Topics include AI bills of materials (BOMs) and least-privilege for AI assistants.

Leverages machine learning for cyber incident attribution and predictive threat analysis, including AI-driven darknet intelligence.