Your question is Data Governance for Regulated AI. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you approach data governance and security when architecting an AI solution for a regulated industry?
Explain how you would establish requirements, classify and control data, secure the ML lifecycle, and enforce appropriate access policies. Address AWS services where relevant, including Amazon SageMaker, AWS Lake Formation, AWS IAM, AWS KMS, Amazon Macie, AWS CloudTrail, Amazon GuardDuty, and Amazon Bedrock. Cover online inference, batch processing, auditability, monitoring, incident response, and controls for training-serving skew and model or feature drift.
