Share via

715-123420 An error occurred. Please reach out to support for additional assistance.

Geoffrey Gunawan 0 Reputation points
2026-04-17T00:24:51.23+00:00

When trying to deploy any model in Microsoft Foundry, i get thrown an error

{ "error": { "code": "715-123420", "message": "An error occurred. Please reach out to support for additional assistance." } }

I'm assuming a support request is required, but i am unable to create it.

Foundry Models
Foundry Models

A catalog of AI models in Microsoft Foundry that you can discover, compare, and deploy using Azure’s built‑in tools for evaluation, fine‑tuning, and inference


2 answers

Sort by: Most helpful
  1. Karnam Venkata Rajeswari 2,390 Reputation points Microsoft External Staff Moderator
    2026-04-27T20:58:54.9466667+00:00

    Hello @Geoffrey Gunawan ,

    Welcome to Microsoft Q&A .Thank you for reaching out to us.

    The 715-123420 indicates a backend issue that requires direct support intervention rather than a configuration change in the portal.

    This type of behavior is commonly associated with platform-level validation controls, which may include automated security or risk-based checks applied at the subscription level. When such controls are triggered, deployment operations can be uniformly restricted, before the model deployment is provisioned.

    For further assistance , a support ticket can be raised following the below suggested steps to involve the backend team to unblock the deployments.Please note that there could be minor UI changes, but the overall process would remain similar-

    1. Please go to the Azure portal > Help + support > New support request.
    2. Then for the Issue type select “Technical” and then under Service select “Azure AI Foundry” (or “AI Services”).
    3. Under Problem type, please choose “Other service issue”
    4. In the description, include: • Error code: 715-123420 , please provide the correlation IDs with a note that "suspect an RTFP/fraud‐block and would like it lifted"
    5. Then please submit the case. The RTFP team will validate your subscription status and remove any automated risk hold.

    The support process enables internal teams to review backend validation systems and determine whether any restriction is in place, followed by appropriate remediation if applicable.

    At this stage, progress depends on backend review by the responsible service teams. Assistance will continue through the support channel to ensure the request is investigated and addressed appropriately.

    Meanwhile we request to please double-check that

    • No policy assignments at the management-group or subscription level are set to Deny on Foundry deployments (see “Control AI model deployment with built-in policies”).
    • No pending compliance or billing holds on your subscription.
    • A paid (not free/trial) subscription is being used—RTFP blocks sometimes target trial accounts.

    Please note to check the following -

    1. Verifying subscription type - Ensure subscription is eligible for AI model deployment workloads. Confirm support for AI service provisioning and model hosting
    2. Validating AI service access approval - Check whether access to AI model deployment services has been granted for the subscription or tenant

    The following references might be helpful , please check them out

    Thank you

    Please 'Upvote'(Thumbs-up) and 'Accept' as answer if the response was helpful. This will be benefitting other community members who face the same issue.

    0 comments No comments

  2. Sucheta Gawade 0 Reputation points MVP
    2026-04-17T04:12:19.9466667+00:00

    That 715-123420 error could be a platform-side block. It can occur when the deployment request is blocked by an automated risk / fraud protection validation, before the model is provisioned.
    Can you try another region? If it’s blocking any model in any region, it’s likely you’ll need to get with Microsoft support to clear the backend block.
    If needed, you can create an Azure support request and select Cognitive Services as the service type. Here is more info on support: https://learn.microsoft.com/en-us/azure/ai-services/cognitive-services-support-options


Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.