Deactivate DSPM for AI - Communication Compliance policy

Nancy Daniel 20 Reputation points
2026-09-14T14:33:09.1233333+00:00

Hello,

It seems that a Communication Compliance policy was created accidentally by me (Purview Admin) while exploring the page DSPM for AI policies recommendations page. Could be one of the "One-Click policies" (DSPM for AI: Unethical behavior in AI Apps). Is there a possibility to deactivate this policy? If so, please guide/provide the steps on how to deactivate the policy.

Thank you for your support!

Best regards,

Nancy

Microsoft Security | Microsoft Purview
0 comments No comments

1 answer

Sort by: Oldest
  1. AI answer

    2026-09-14T14:33:21.8633333+00:00

    Yes. A Communication Compliance policy can be deactivated by pausing it.

    Use these steps in Microsoft Purview:

    1. Go to the Communication Compliance solution.
    2. Open the Policy page.
    3. Select the policy that was created, such as the DSPM for AI one-click policy.
    4. On the actions toolbar, select Pause policy.
    5. In the Pause policy pane, confirm by selecting Pause.

    Important details:

    • Pausing a policy stops inspection and alert generation for all user message conditions in that policy while it is paused.
    • Existing alerts and messages created before pausing remain available for investigation, review, and remediation.
    • It can take up to 24 hours for the policy to be fully paused.
    • The status can show Pausing first, and then Paused when complete.
    • To do this, membership in the Communication Compliance Admins role group is required.

    If the policy appears with a prefix such as DSPM for AI, that indicates it was created in Data Security Posture Management for AI, but review and remediation actions are handled from the Communication Compliance solution.


    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

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.