What the Copilot policy location is, which template offers it and the controls it provides.
From Ultra Transcenders AB-650 by Tony Rough (coming November 2026)
The Microsoft 365 Copilot and Copilot Chat policy location lets DLP control what Copilot, Copilot Chat and Cowork can use or process. It is available only in the Custom policy template, and choosing it disables every other location in that policy. Updates can take up to four hours to reach Copilot.
| Condition | Action | Effect |
|---|---|---|
| Content contains > Sensitive information types | Prevent Copilot from processing content > Performing Web Searches | External web search is blocked for that prompt; the answer is still grounded in internal sources |
| Content contains > Sensitive information types | Prevent Copilot from processing content > Processing prompts | No response is returned, and the prompt is not used for internal or web search (preview) |
| Content contains > Sensitivity labels | Prevent Copilot from processing content | Labelled files and emails are not used in the response, although they can still appear in citations |
| Email is received from > External users | Prevent Copilot from processing content | Emails from outside the accepted domains are excluded from grounding, summarising and citation (preview) |
Key points:
https://) can replace it.Roles that can create or edit these policies include Microsoft Entra AI Admin, Purview Data Security AI Admin (which cannot read prompts and responses), Purview Compliance Administrator, Purview Information Protection Admin, Purview Security Administrator and Entra Global Administrator.
The default policy, named Default DLP policy - Protect sensitive M365 Copilot interactions, starts in simulation mode with policy tips shown. It logs matches and sends a low-severity alert email if one is configured, but it blocks nothing until the state is changed to turn it on. Microsoft recommends turning on incident reports and adding the security team as recipients.
Common trap: Leaving the default Copilot policy in place and assuming prompts containing credit card numbers are blocked - it starts in simulation mode, so it only records matches until an administrator enforces it.
This note is one section of Ultra Transcenders AB-650: Administering Microsoft 365 and AI Services, an independent study guide that explains every topic the exam covers by technology, with comparison tables, diagrams and the common traps, plus a glossary linked to Microsoft Learn.
Due on Amazon in November 2026, in Kindle and paperback editions.
About the book · AB-650 terms in the glossary · All AB-650 study notes
Which groups can carry licences, who can assign them and how membership drives access.
How administrative units scope roles to a subset of users, groups and devices.
How built-in and custom authentication strengths define which method combinations satisfy a policy.
How a Conditional Access policy combines assignments, conditions and access controls.
What the harmless phishing simulations are, and the licence and permissions needed to run them.
How SharePoint data access governance reports and Purview data risk assessments help to find oversharing.
The three audience options for when tenant users receive new features, and where Frontier fits.
Where Copilot issues are tracked: Service health for incidents and Message center for changes.