‹ All AI-300 questions

AI-300 Practice Question 93: Online Endpoint Entra Auth

Microsoft Machine Learning Operations Engineer Associate (AI-300) · Test 10 · question 3 of 10

Implement machine learning model lifecycle and operationshard

Client applications call a managed online endpoint by using managed identities. Security forbids shared keys, and wants the scoring permission to be checked by Azure RBAC and granted or revoked per identity. Which authentication mode should the endpoint use?

Select one answer
Show answer and explanation

Answer:

  • B. auth_mode: aad_token, with a scoring role assigned to each client identity

Managed online endpoints support key, Azure Machine Learning token, and Microsoft Entra token (aad_token) authentication. Only aad_token is checked against an RBAC scoring permission at the endpoint, so access can be granted or revoked per managed identity with role assignments (option B).

- A is wrong because the keys are still shared secrets that do not carry the caller's identity, and access cannot be managed per identity through RBAC. - C is wrong because an aml_token is issued by the workspace to anyone holding the token permission, and the token itself is not checked against an RBAC scoring permission when the endpoint is called. - D is wrong because a private endpoint restricts network paths, but the callers would still authenticate with a shared key.

Reference: Authenticate clients for online endpoints

Practice Test 10 ›

Study guides for AI-300