The correct answer is Input validation.

Input validation ensures that values received from clients match the expected data types formats and constraints before the application processes them. It is implemented with allow lists type checks length limits and pattern matching on the server so that malformed or malicious data is rejected early and does not reach sensitive logic.

Cloud Audit Logging is a logging and audit capability and it does not itself verify or enforce that client inputs conform to expected types or formats. It is useful for monitoring and forensics after the fact but it does not perform validation.

Error and exception handling concerns how an application responds to runtime failures and it helps maintain stability and avoid information leaks. It does not validate incoming data and therefore cannot guarantee that values match expected types or formats.

Output sanitization is focused on cleaning or encoding data before rendering it to users to prevent cross site scripting and other injection attacks. It does not replace validation because it does not enforce type or format constraints before data is processed.

Full AWS Practitioner Certification Question