Data Security and Privacy for GenAI

Keeping the data honest while the model uses it: VPC endpoints and IAM for private model access, Lake Formation permissions, PII detection with Comprehend and Macie, masking and anonymization, and retention design.

Content current as of 2026-10.

Lessons

1. Private paths to models
2. PII in GenAI pipelines
3. Retention, lifecycle, and the right to be forgotten

Sources

• AIP-C01 exam guide — Content Domain 3 (task 3.2)
• Protect your data using Amazon VPC and AWS PrivateLink (Bedrock)
• Amazon Comprehend — personally identifiable information (PII)
• Amazon Bedrock — model invocation logging