AWS Certification Exam AI Practitioner (Practice Questions) set12
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 52 / 65
Which mechanism should be implemented to later verify whether the model's outputs have been tampered with or false instructions have been given?
①Require the user to manually confirm all outputs.
②Encrypt the output so that it cannot be decrypted.
③Completely erase the output contents each time to make auditing impossible.
④Publish logs externally without restrictions and make them editable by anyone.
⑤Implement signatures and traceability on outputs to enable detection and tracking of modifications.