5.2Standard
A public chat receives huge or unsafe inputs.
Reduce cost and attack surface.
Choose two controls.
× Incorrect
Validate length, format, encoding, file type, rate, and dangerous instructions, with bounded errors, quarantine, review, and minimized logs.
Detailed explanation
CorrectInspect length, format, encoding, file type, rate, and dangerous instructions
Oversized and some attack inputs are reduced.
IncorrectAccept unlimited input and send everything to the model
Cost and attack surface grow.
CorrectReturn bounded errors, limits, quarantine, or human review and minimize rejection logs
A safe alternate path exists.
IncorrectStore rejected input in plaintext without limits
Sensitive data remains.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 5.2・2.3の入力検証、インジェクション、レート制限を確認する。
Official references
What to confirm
入力を無条件にモデルへ渡さず、拒否・隔離・ログ保護を設計できる。
Key points
- Input validation
- Limit
- Quarantine
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
- Compare the definition with the AIF-C01 objective and the service documentation before applying it.
- Keep the concept separate from a particular answer choice or implementation step.
Foundation review
Input validation
Checking incoming data against expected type, size, format, and policy constraints.
Isolation
Separating untrusted processing from sensitive resources and trusted operations.
Question IDAWS-AIF-503
EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
Last technical review2026-08-20
Report an error or rights issue