You are working with developers to secure custom training jobs running on Vertex AI. For compliance reasons, all supported data types must be encrypted by key materials that reside in the Europe…

Question

Options

• AEncrypt the code, training data, and metadata with Google default encryption. Use customer-
• BEncrypt the code, training data, metadata, and exported trained models with customer-managed
• CEncrypt the code, training data, and exported trained models with customer-managed encryption
• DEncrypt the code, training data, and metadata with Google default encryption. Implement an

• A
  5% (3)
• B
  11% (6)
• C
  82% (45)
• D
  2% (1)

Explanation

Customer-Managed Encryption Keys (CMEK) allow organizations to control the key material used to encrypt data in Google Cloud services, with keys stored in Cloud KMS in a specified region (Europe, in this case). For Vertex AI training jobs, CMEK can be applied to code, training data, and exported trained models. The question specifies 'all supported data types' - metadata encryption is not a supported CMEK target in Vertex AI, so Option B (which includes metadata) is incorrect. Option C correctly targets only the supported types. Options A and D use Google default encryption for some components, which means key material remains under Google's control rather than the organization's, violating the compliance requirement. CMEK does not impact training operations as encryption/decryption is handled transparently by the KMS integration.

No community discussion yet for this question.