Test goal hijacking, tool misuse, memory poisoning, multi-agent attacks, and cascading failures in safe AI sandboxes
Identify and reduce prompt injection, data leakage, supply chain attacks, and agent security risks with practical labs.
Learn Jev with Python: build model routing, AI guardrails, citation checks, benchmarks, and safe decision workflows
Apply AI to nuclear monitoring and digital twins; explore fusion, SMRs, reactor safety, and data center energy.