Can a Copilot Agent Use a Shared Mailbox as a Knowledge Source?

(1) Share

Report

Posted on by CU04060744-0

Hi everyone,

I would like to understand whether it is possible to use a shared mailbox as a knowledge source for an agent created in either Microsoft 365 Copilot or Copilot Studio.

My use case is as follows:

A shared mailbox contains several years of email correspondence between staff and external parties.

Many of the questions received are repeated over time, and the answers can often be found in previous email conversations.

I would like the agent to search the historical emails in the shared mailbox, identify relevant past responses, and use that information to generate an answer to the user's query.

Specifically, I would like to know:

Can a shared mailbox be configured as a knowledge source for a Copilot agent?

Can the agent search email content (including historical email threads) within the shared mailbox?

If direct integration with a shared mailbox is not supported, what would be the recommended design pattern to allow the agent to leverage past email correspondence as a knowledge base?

Any guidance or best practices would be greatly appreciated.

Thank you.

Categories:

Autonomous agents

All responses (5)

Answers (0)

Sort by

Suggested answer

Haque 4,097 Super User 2026 Season 2 on at

Like
a
(1)

Report
Copy link

Link copied!

Hi @CU04060744-0,

Yes its possible, please refere to section "Email Knowledge - Handle With Care" in this article/blog.

Also, we have resolved thread on the same.

Please let me know if these clue help.

I am sure some clues I tried to give. If these clues help to resolve the issue brought you by here, please don't forget to check the box Does this answer your question? At the same time, I am pretty sure you have liked the response!
Suggested answer

11manish 4,388 Super User 2026 Season 2 on at

Like
a
(0)

Report
Copy link

Link copied!
At present:

A shared mailbox cannot be configured as a native knowledge source for Microsoft 365 Copilot or Copilot Studio.

Agents cannot automatically perform semantic searches across historical shared mailbox conversations as they do with SharePoint-based knowledge.

The recommended approach is to move or synchronize valuable email content into a supported knowledge repository such as SharePoint, Dataverse, or Azure AI Search, depending on the scale and sophistication of your solution.

For large volumes of historical email correspondence, a RAG architecture using Microsoft Graph to ingest emails and Azure AI Search for indexing and retrieval is generally considered the most scalable and enterprise-ready design.
Suggested answer

Valantis 7,266 Super User 2026 Season 2 on at

Like
a
(1)

Report
Copy link

Link copied!

Hi @CU04060744-0,

Confirmed from Microsoft docs: a shared mailbox cannot be added as a native knowledge source in either Copilot Studio or M365 Copilot Agent Builder.

<cite index="62-1">You can't scope email knowledge. When you add email, the agent uses all email in your mailbox as knowledge. Users that you share the agent with don't have access to your email as knowledge.</cite> So even the personal mailbox option only works for the creator's own inbox, not for a shared mailbox, and doesn't extend to other users.

For your use case (shared mailbox with years of historical correspondence as a knowledge base), the recommended pattern is:

1. Export historical emails to SharePoint: use Power Automate to sync emails from the shared mailbox to a SharePoint document library (as .eml files, HTML exports, or text files per email). Then add that SharePoint library as a knowledge source in Copilot Studio.

2. For ongoing emails: set up a flow triggered on new emails to the shared mailbox that saves them to the same SharePoint library, keeping the knowledge base current.

This gives you full semantic search over historical threads, works with all users (SharePoint permissions control access), and is the most supported pattern for this scenario at the moment.

Best regards,

Valantis ✅ If this helped solve your issue, please Accept as Solution so others can find it quickly.

❤️ If it didn’t fully solve it but was still useful, please click “Yes” on “Was this reply helpful?” or leave a Like :).

🏷️ For follow-ups @Valantis.

📝 https://valantisond365.com/ 💼 LinkedIn ▶️ YouTube
Suggested answer

Ashlesha-MSFT Microsoft Employee on at

Like
a
(1)

Report
Copy link

Link copied!

Hi CU04060744-0,

I validated this in our tenant and can confirm the following:

- In the current Agent Builder UI path we tested, the Knowledge section only exposes website-based knowledge input (URL/scoped web options). A shared mailbox picker is not exposed there.
- So, for this UI flow, a shared mailbox cannot be added directly as a native knowledge source.
- If your goal is to use shared mailbox history, the practical supported pattern is to ingest/sync mailbox content into a supported knowledge store (for example SharePoint, Dataverse, or Azure AI Search), then ground the agent on that store.
- For declarative/manifest-based builds (advanced path), email capability options may differ from the simplified UI, but in this UI experience the mailbox option is not available.

Recommended design for your scenario:
1. Sync historical and new shared mailbox emails into SharePoint (or another supported indexed store).
2. Add that repository as agent knowledge.
3. Keep permissions aligned so retrieval respects access controls.

References:
- https://learn.microsoft.com/en-us/microsoft-copilot-studio/knowledge-copilot-studio
- https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/agent-builder-add-knowledge
- https://learn.microsoft.com/en-us/microsoft-365-copilot/extensibility/knowledge-sources
Valantis 7,266 Super User 2026 Season 2 on at

Like
a
(0)

Report
Copy link

Link copied!

Hi @CU04060744-0,

Just wanted to check in and see if everything is working now. If you still need any help, feel free to let me know.

Also, if the issue is resolved, it would be great if you could mark the answer as solved so others with the same question can find it easily.

Thanks and have a great day!