A new AI feature infers personal information.
Assess impact before launch.
Choose two practices.
Map data, purpose, subjects, inferences, sharing, retention, and rights, then review reidentification, inference, transfer, access, and reuse risks.
Detailed explanation
Processing is visible.
Processing is visible.
Prevention is possible.
Prevention is possible.
Inferences and combinations still matter.
Inferences and combinations still matter.
Prevention is delayed.
Prevention is delayed.
Try it yourself
An example you can run in a temporary verification environment.
AWS公式AIF-C01 Domain 5.1のプライバシー影響、個人データ、リスク評価を確認する。Expected result
直接識別子だけでなく推論・組合せによるプライバシー影響を事前に評価できる。Key points
- PIA
- Inference
- Reidentification
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Command output formatting can vary slightly by distribution or tool version.
- Run the example in a temporary directory or process when possible.
Foundation review
Read the scope first
Check whether the command acts on the current shell, a new process, an existing process, or a file.
Verify the observable result
Use the supplied command and compare the output with the expected result.