Azure OpenAI Abuse Monitoring Opt-out

Marek Pavelka 100 Reputation points
2025-07-30T14:28:27.95+00:00

Hello,

I would like to ask if there is a way to get exception from Abuse Monitoring in Azure OpenAI? We have tried multiple times to fill in the form (https://customervoice.microsoft.com/Pages/ResponsePage.aspx?id=v4j5cvGGr0GRqy180BHbR7en2Ais5pxKtso_Pz4b1_xUOE9MUTFMUlpBNk5IQlZWWkcyUEpWWEhGOCQlQCN0PWcu) but always got rejected for reason not being managed customer.

Our use case is that our customers are lawyers and they send to our service their contracts they have for their clients we process the contracts with Azure OpenAI LLMs. Hence the customers require us to ensure their data privacy and are not happy about the 30 days retention period for the abuse monitoring purpose.

Hence I would like to ask if there is some other way how we could apply to get the exception from abuse monitoring?

Kind regards,

Marek Pavelka

Azure OpenAI in Foundry Models

1 answer

Sort by: Most helpful
  1. Pavankumar Purilla 11,655 Reputation points Moderator
    2025-07-30T15:33:15.4666667+00:00

    Hi Marek Pavelka,

    To request an exception for abuse monitoring, Microsoft has specific eligibility criteria—this process is primarily designed for managed customers who have a dedicated Microsoft account team. Since your requests have been rejected for "not being a managed customer," there is currently no direct way to bypass this requirement.

    If you believe your organization should qualify for managed status, we recommend reaching out to your Microsoft account team or contacting Azure Sales through the Microsoft Azure Contact Page to discuss your options. If your organization becomes a managed customer, you can reapply for the exception using the Abuse Monitoring Opt-Out Request form.

    Without managed customer status, the 30-day retention period for abuse monitoring will remain in place.

    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.