Placeholder risk description for privacy and security failures in AI systems.
Treatment placeholder: data minimization, access control, red-teaming, and model/security hardening.
Back to CatalogMIT Risk Domain
Privacy & Security
Leakage of sensitive information and vulnerability of AI systems to attack.
Turn this into a program
AI Risk Guy maps regulations, risks, and frameworks into one governance program — so you manage obligations, not documents.
Talk to us