Subject: GPT-5-mini Deployment Blocked – Unusual Activity Detected

Question

Subject: GPT-5-mini Deployment Blocked – Unusual Activity Detected

Ahsan Nawaz 20

Hello Microsoft Azure Support,

I am trying to deploy the GPT-5-mini model in Microsoft Foundry using the Global Standard deployment type in the North Central US region.

However, deployment repeatedly fails with the following error:

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

I have verified model availability and reduced the token-per-minute limit to 10,000, but the issue persists.

Could you please review my subscription and resource for any security or deployment restrictions and advise how to resolve this issue?

Thank you.

Sairam Reddy Rajamuri 0 Reputation points Microsoft External Staff Moderator

2026-10-22T07:38:58.1+00:00
Hello Ahsan,

The message indicates that the deployment request requires an account- or subscription-specific review. Because the error does not expose the exact validation reason, it is not possible to determine from Microsoft Q&A whether a security or deployment restriction has been applied.

Reducing the token-per-minute allocation does not address the condition identified by this error. Since you have already confirmed model availability and the deployment continues to fail at 10,000 TPM, please create an Azure support request from the affected resource:

In the Azure portal, open the affected Microsoft Foundry resource.

Under Help, select Support** + **Troubleshooting.

Provide the error message and complete any offered diagnostics.

If the issue remains unresolved, select Create** a support **request.

Include the following information:

Subscription and tenant IDs

Resource name and full resource ID

Model and model version

Region and deployment type

Exact UTC time of a recent failed attempt

Complete error details

Request, tracking, or correlation ID from the failed Activity Log operation

Please provide subscription, tenant, resource, and correlation identifiers only through the secure support channel, not in this public thread.

Microsoft Support can use these details to investigate the account-specific validation and advise whether any action is required. The public error message alone does not confirm the type or cause of the restriction.

Thank You!
Sairam Reddy Rajamuri 0 Reputation points Microsoft External Staff Moderator

2026-10-23T04:57:15.6933333+00:00

Hello Ahsan,

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank You!

Answer accepted by question author

Senthil kumar 3,000

Hi @Ahsan Nawaz

Verify:

The subscription is active and not disabled.
The Azure OpenAI / Foundry resource was created successfully.
The subscription has quota assigned for GPT-5-mini in the target region.
No Azure Policy is denying AI resource operations.
The account has appropriate permissions (Owner, Contributor, or Azure AI role assignments).

Thanks.

0 comments

1 additional answer

Your answer

Sairam Reddy Rajamuri 0 Reputation points Microsoft External Staff Moderator

2026-10-22T07:38:58.1+00:00

Hello Ahsan,

The message indicates that the deployment request requires an account- or subscription-specific review. Because the error does not expose the exact validation reason, it is not possible to determine from Microsoft Q&A whether a security or deployment restriction has been applied.

Reducing the token-per-minute allocation does not address the condition identified by this error. Since you have already confirmed model availability and the deployment continues to fail at 10,000 TPM, please create an Azure support request from the affected resource:

In the Azure portal, open the affected Microsoft Foundry resource.

Under Help, select Support** + **Troubleshooting.

Provide the error message and complete any offered diagnostics.

If the issue remains unresolved, select Create** a support **request.

Include the following information:

Subscription and tenant IDs

Resource name and full resource ID

Model and model version

Region and deployment type

Exact UTC time of a recent failed attempt

Complete error details

Request, tracking, or correlation ID from the failed Activity Log operation

Please provide subscription, tenant, resource, and correlation identifiers only through the secure support channel, not in this public thread.

Microsoft Support can use these details to investigate the account-specific validation and advise whether any action is required. The public error message alone does not confirm the type or cause of the restriction.

Thank You!
Sairam Reddy Rajamuri 0 Reputation points Microsoft External Staff Moderator

2026-10-23T04:57:15.6933333+00:00

Hello Ahsan,

Did you get any chance to review the above response. Do let me know if you have any further queries.

Thank You!

Answer 1

Request a Microsoft account/subscription-level review for the deployment block, because reducing TPM does not clear the “unusual activity” validation and comparable 715-123420 cases were resolved only after Microsoft adjusted or unblocked the subscription.^1^

Submit the review request

In the Azure portal, open Help + support and select Create a support request.
Describe the issue as:

Microsoft Foundry gpt-5-mini Global Standard deployment blocked by “Our system has detected this request as unusual activity for your account.” Request account/subscription security-validation review.
Select the affected subscription and Foundry resource.
If available, choose Subscription management or Billing/subscription rather than quota increase. Azure provides subscription-management support independently of technical support plans.^2^
Include:
- Exact error message and error code, particularly 715-123420 if displayed
- Subscription ID and tenant ID
- Foundry resource and resource-group names
- Model and version: gpt-5-mini
- Deployment type: Global Standard
- Region: North Central US
- Requested capacity: 10,000 TPM
- Failure timestamp in UTC
- Correlation ID
- Confirmation that model availability and quota were already verified

Do not post subscription IDs, tenant IDs, or correlation details publicly; provide them only through the private support request.

The account used to create the request must have Owner, Contributor, Support Request Contributor, or equivalent Microsoft.Support/* permission at subscription scope.^2^

Collect the correlation ID

In the Azure portal:

Open the resource group containing the Foundry resource.
Select Deployments or Activity log.
Open the failed Microsoft.CognitiveServices/accounts/deployments/write operation.
Copy its Correlation ID, timestamp, and complete error details.

Azure Support uses deployment correlation IDs to locate related operations.^3^

What has already been ruled out

No further TPM reduction or regional availability check is warranted from the information provided:

gpt-5-mini version 2025-08-07 supports Global Standard deployment in northcentralus.^4^
Similar failures occurred despite available quota and were rejected during an additional validation stage before normal provisioning.^5^
Other matching cases were resolved by backend adjustment or subscription unblocking.^6^

Avoid repeatedly deleting and recreating the Foundry resource or deployment. If the subscription is new and automation is being used, submit the first deployment manually through Foundry after the restriction is released; this was the specific guidance in a matching first-deployment case.^1^

If the portal does not expose a usable support-request path, add these details to the Microsoft Q&A case and request private moderator-assisted routing. Do not place account identifiers in the public post.

References