A hands-on, live 4-session masterclass on breaking and fortifying AI systems: master adversarial prompt injection, jailbreaks with garak and PyRIT, enterprise guardrails, EU AI Act compliance, and MATLAB formal neural network verification.
EU AI Act Annex IV documentation, corporate AUP drafting, and training dataset copyright/GPL clean-room audits.
Direct and indirect prompt injection, multi-turn Crescendo jailbreaks, and automated fuzzing with garak and PyRIT.
Implementing NVIDIA NeMo Guardrails, Meta Llama Guard 3, and real-time Presidio PII masking pipelines.
Reachability analysis in Deep Learning Toolbox, Simulink barrier supervisors, and ISO 26262 ASIL-D certification.
Live Interactive Zoom Sessions with Q&A