AWS Certified AI Practitioner
Securing AI Workloads with AWS Services
The AWS controls the exam names for AI systems: the shared responsibility model, IAM, KMS encryption, PrivateLink, Macie, GuardDuty, Inspector, AgentCore Identity, and invocation logging.
Intermediate 24 minutes 6 Learning Objectives
- Apply the AWS shared responsibility model to a managed AI service such as Amazon Bedrock
- Describe how IAM roles and policies control access to models, data, and agent actions
- Identify which Amazon Bedrock resources support customer managed KMS keys
- Explain what AWS PrivateLink changes about the network path to an AI service
- Match a security requirement to the right service among Macie, GuardDuty, Inspector, and CloudTrail
- Distinguish CloudTrail from Amazon Bedrock model invocation logging as audit evidence