INFRASTRUCTURE LEARNINGINFRASTRUCTURE LEARNING
インフラ大全インフラ大全IT INFRASTRUCTURE / LEARNING
Infrastructure learning guide
Find a lesson or lab/

Esc to close · Search terms stay on your device

LEARNING日本語

Topic lessonTopic: Amazon Bedrock readiness, privacy and evaluation

Protect Amazon Bedrock prompts and invocation logs

Check data classification, model processing and retained evidence individually. Use a synthetic database operations task to test what is collected and who can read it.

What you will learn

Use a synthetic database operations task to test what is collected and who can read it.

Prerequisite

This exercise uses fictional evidence and makes no changes in a live environment.

Follow data from input to retention

Model processing and operator-configured invocation logging use separate controls. Review endpoint, Region, destinations and readers.

From classification through access
From classification through access1: Input — Confirm data scope — Process. 2: Process — Verify configuration — Log. 3: Log — Reconcile audit evidence — Read123InputProcessLogReadFrom classification through access1: Input — Confirm data scope — Process. 2: Process — Verify configuration — Log. 3: Log — Reconcile audit evidence — Read123InputProcessLogRead
  1. Input → ProcessConfirm data scope
  2. Process → LogVerify configuration
  3. Log → ReadReconcile audit evidence
Input
Classify and minimize
Process
Fix the invocation path
Log
Identify what is collected
Read
Assign readers and retention

Summarize a synthetic database runbook

Operators need limited audit visibility. The fixture contains no confidential values.
ControlDecisionTest
InputSynthetic operations guideNo personal or secret values
OutputSummary with test observationsDo not duplicate sensitive values in records
Invocation logVerify collection scopeFind the intended synthetic record
Retention and readersRestrict to the review roleCheck expiry and access boundaries

Use a synthetic event to check collection and access without introducing sensitive data.

Evidence for the decision

Define the data responsibility Decide necessity, classification, processing location and the owner of each setting.

Check collection scope Match invocation mechanism, logging settings, destinations and recorded fields.

Protect stored content Verify access, retention, encryption and unintended sharing with a synthetic test.

Check your decision

Choose an option to see the result and explanation.

Can you skip logging checks because the input is synthetic?
Read the answer and reason

Answer: Match the invocation and recorded fields · Exercise operational controls before deployment.

Do logging settings collect every invocation the same way?
Read the answer and reason

Answer: Check the selected path and synthetic record · Check the selected mechanism and applicable settings.

How do you verify log access?
Read the answer and reason

Answer: Required read succeeds and unrelated access fails · Use intended identities and scope.

Check, then explain in your own words

  1. Classify the input.
  2. Review invocation and collection.
  3. Verify storage access and expiry.
Check your understanding: Does processing imply that the entire prompt appears in your audit storage?

Verify the invocation mechanism, logging configuration and specific fields.

Processing and collection are separate functions.

Sources and conditions checked

The official references used in this lesson are collected here with their scope, conditions and corresponding sections. Independent expert review and live AWS comparison have not been performed.