Until yesterday we've seen "responsible AI" — making AI fair and trustworthy. Today we shift to technical defense. AI Security — Who can access models and data, how is data protected, how is personal information guarded. As a cloud certification, AIF-C01 asks about security fundamentals. There's no separate AI-only security technology; the core is applying AWS's standard security principles in an AI/ML context.
Today's five pillars: least privilege IAM, data encryption (at-rest/in-transit), PII protection, PrivateLink, and the foundation of all security — the shared responsibility model.