azd up fails - "715-123420 unusual activity" on Microsoft.CognitiveServices/accounts preflight

Alessia Saggio 0 Reputation points
2026-10-15T09:39:08.9533333+00:00

Hello all,

I'm deploying an Azure AI Foundry template (Microsoft's mslearn-genaiops azd template) and every provision fails at preflight validation.

 

Error:

InvalidTemplateDeployment — Microsoft.CognitiveServices/accounts (2025-06-01) reported preflight validation errors:

715-123420: "Our system has detected this request as unusual activity for your account. If you are confident this is in error, please contact support."

 

What happens:

  • The resource group and base resources (Log Analytics, App Insights) provision fine.
  • It consistently fails when creating the CognitiveServices / Foundry account.
  • Same error from azd up, azd provision, and azd provision --preview.
  • The error occurs at PREFLIGHT, before any account or model deployment is actually created.

 

What I've already ruled out:

  • Not quota: az cognitiveservices usage list shows current usage at 0 against non-zero limits.
  • Not SKU/region availability: az cognitiveservices account list-skus --kind AIServices shows S0 is available in my region.
  • Not a transient conflict: the block has persisted 18+ hours across many attempts and changing deployment names.
  • Not a template bug: it's an unmodified official Microsoft azd template, and the base resources deploy successfully.

 

This looks like an account-level anti-abuse / "unusual activity" flag rather than a configuration or quota problem. I'm attending the AI-300 course and I suspect this is arising from spinning-up and deleting resources quite frequently from following the exercises. But of course just a hypothesis.

 

Questions:

  1. Has anyone had this 715-123420 "unusual activity" flag on a personal subscription, and how did you get it removed?
  2. Is contacting support the only way to clear it, or is there a self-service unblock (e.g. identity verification, adding a payment method, subscription upgrade)?
  3. How long did it take to be lifted once reported?

 

Thanks in advance.

Microsoft Foundry
Microsoft Foundry

A unified Azure platform for creating and managing AI models, agents, and applications with built‑in enterprise security, monitoring, and governance


2 answers

Sort by: Most helpful
  1. Alessia Saggio 0 Reputation points
    2026-10-15T10:21:47.4633333+00:00

    Thanks for your reply @Senthil kumar ,

    Was hoping there was a faster solution to this, but alas. I contacted support.

    Was this answer helpful?

    0 comments No comments

  2. Senthil kumar 2,585 Reputation points
    2026-10-15T09:52:43.5433333+00:00

    Hi @Alessia Saggio

    1. Has anyone seen this and how was it removed? Yes. Multiple users have reported the 715-123420 error. Microsoft moderators indicate that the flag is applied at the subscription/account level and generally requires investigation by Microsoft's internal fraud/security team. Users reported working through Microsoft Support rather than resolving it through resource configuration changes.

    2. Is there a self-service unblock? I found Microsoft guidance stating that this type of block requires review by the appropriate internal team and that users should contact support for investigation/escalation. The sources do not mention a self-service method such as identity verification, adding a payment method, upgrading the subscription, or changing settings to clear the flag.

    3. How long does it take to be lifted? I could not find an official timeframe in the sources. Microsoft Q&A responses indicate that the case must be reviewed by internal teams, but they do not provide a guaranteed duration for removal of the flag.

    Practical takeaway:

    • Recreating resources, changing regions, or repeatedly retrying deployments is generally not reported as a fix.
    • Open a support request and provide the subscription ID, tenant ID, affected resource details, and the exact error code (715-123420). Microsoft moderators consistently direct users toward support/internal review.

    Thanks.

    Was this answer helpful?

    0 comments No comments

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.