The correct answer is: Update the S3 Lifecycle configuration to also delete previous object versions in addition to current versions.

S3 Versioning keeps all versions of an object, including deleted or replaced versions. If the lifecycle policy only deletes current versions, older noncurrent versions will accumulate over time. To reduce storage costs and enforce retention properly, the lifecycle policy must also include a rule to delete noncurrent (previous) versions after the retention period. This ensures that all object versions older than 3 years are removed automatically and efficiently.

Why the other options are incorrect:

Setting CloudTrail to expire objects is not a valid feature. CloudTrail does not have a built-in expiration or deletion mechanism for S3 objects; retention is managed through S3 lifecycle policies.

Creating a Lambda function to manually delete old objects is technically feasible but introduces unnecessary complexity and ongoing cost. S3 lifecycle policies are the native, low-cost, and automated way to manage object expiration.

Changing the object ownership to the parent account does not affect object versioning or lifecycle management. S3 lifecycle policies apply regardless of which account owns the object, assuming proper permissions are in place.

Why this question is important: It tests your understanding of how S3 Versioning and Lifecycle policies interactespecially in long-term logging or archival scenarios where storage costs and automated retention policies are critical.

Tip: Always check whether lifecycle rules are configured for both current and previous object versions when using versioned buckets, especially for log storage like CloudTrail.

Full AWS Practitioner Certification Question