Azure ML won't create Workspace, giving Interservererror

Hello Office,

Welcome to the Microsoft Q&A and thank you for posting your questions here.

I understand that your Azure ML won't create Workspace, and it's giving Interservererror.

Though, there is "general Internal Server Error" (500) is usually caused by a problem on Azure's side (like server problems or outages). But in your case, Azure ML catches that you are not properly authenticated (wrong account type, broken token, missing claims, wrong login method), but (badly) wraps it in a 500 error instead of returning a cleaner 401. The error message specifically says:

Unauthorized

InvalidAuthenticationToken

Missing claims (oid, puid, altsecid)

This is NOT a general server error but it's an authentication problem with your access token before Azure ML can even start doing anything.

You can resolve this by the following steps:

• If you're using a personal Microsoft account, you need to switch to an Azure Active Directory (Azure AD) account because personal Microsoft accounts aren't supported for Azure Machine Learning. You might have to create an Azure AD user or invite yourself as a guest into the tenant.
• Make sure the service principal is properly set up in Azure AD and that the app has the necessary permissions. When you're authenticating, request the token for the correct audience, such as https://management.azure.com/.default.
• When you log in using az login, ensure it shows an account type of "User" or "ServicePrincipal" linked to Azure AD. az login az account show
• If you're doing this in a script or SDK, make sure you're using the DefaultAzureCredential or the correct login method.

Relevant documentation

Using Service principal

Manage user roles

I hope this is helpful! Do not hesitate to let me know if you have any other questions or clarifications.

Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful.