To make business data available to Microsoft 365 Copilot, deploy a Microsoft 365 Copilot connector: choose a synced connector to index external content on a recurring schedule, or a federated connector to query a compatible source at prompt time without indexing it. Before setup, confirm licensing, the source’s identity and query requirements, and who is allowed to see the data.
What connecting business data to Copilot means
This guide concerns organizational data and Microsoft 365 Copilot connectors. A connector makes information from an external system available to supported Microsoft 365 experiences; it does not mean Copilot can automatically access every system or every file in your organization. The source, connector pattern, licensing, identity setup, and permissions determine what can be found and by whom.
![The Microsoft Copilot Bible for Beginners & Busy Professionals: [5 in 1] How to Boost Productivity, Save Hours Every Week, and Automate Your Work with AI](https://m.media-amazon.com/images/I/41tTEg9fueL._SL160_.jpg)
![The Microsoft Copilot Bible: [3 in 1] A Beginner's Guide to Harness AI in Office 365 to Automate Workflows, Generate Content, Analyze Data, and 10x Your Productivity](https://m.media-amazon.com/images/I/41+5Qv4pd2L._SL160_.jpg)
As an Amazon Associate I earn from qualifying purchases.
Microsoft’s connector overview reported more than 100 available connectors on June 18, 2026, including examples for Azure services, Box, Confluence, Google services, MediaWiki, Salesforce, and ServiceNow. That is a dated catalog statement, not a guarantee that a particular connector is currently available or suitable for your tenant. Check the current Microsoft documentation and the deployment guide for your chosen source.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallChoose between synced, federated, and custom paths
| Path | How it works | Best fit and main trade-off |
|---|---|---|
| Synced connector | Copies and indexes content from an external data source into the Microsoft 365 substrate on a recurring schedule. | Use when indexed discovery and Copilot grounding are the goal. Content is ingested and updated through sync, so plan for source credentials, permissions, and freshness expectations. |
| Federated connector | Queries a compatible external source at runtime without indexing its content; the data stays in that source. | Use when the source should remain in place and can support the required live query and identity patterns. Each user querying the source needs eligible licensing. |
| Copilot connectors API | Lets an organization build a custom ingestion and data-modeling path for content not covered by a suitable built-in connector. | Consider when the built-in choices or schemas do not meet the need. It requires development and data modeling; API-ingested items consume item quota. |
| SharePoint Server connector | Uses a Microsoft Graph connector agent installed and registered to the Microsoft 365 tenant. | Investigate for content in an on-premises SharePoint Server farm. It requires agent, environment, authentication, and permission preparation. |
Compare the options against the source’s freshness requirements, user and service identity support, permission fidelity, license eligibility, implementation effort, and ability to support the required query or ingestion pattern. A connector that can technically reach a source is not necessarily the right choice if its identity model cannot enforce the access rules your organization needs.
Check licensing before planning deployment
Microsoft’s current connector prerequisites distinguish between access for Microsoft Search and grounding for Copilot. The matrix below reflects the eligibility described in that guidance; verify the entitlement assigned to your tenant and the users who will query the data before implementation.
| License configuration | Synced connectors | Federated connectors |
|---|---|---|
| Microsoft 365 alone | Supported for Microsoft Search; not for Copilot grounding. | Not listed as supported in the prerequisites matrix. |
| Microsoft 365 plus the Copilot add-on | Supports Copilot grounding. | Supported; each user querying the source needs a Microsoft 365 Copilot add-on license. |
| Microsoft 365 E7 | Listed as supporting synced connectors. | Listed as supporting federated connectors. |
For federated access, the source and Microsoft Entra ID must be configured for OAuth-based identity, and the source must support the required authentication and query patterns. Licensing should be checked for the actual users who will query the source, not only for the administrator who sets up the connector.
Prepare the source, administrator, and access model
An AI administrator deploys connectors in the Microsoft 365 admin center. A prebuilt synced connector may also require an administrator to access the external service. For a synced setup, have the source credentials, source-specific environment details, and permission information ready. For federation, coordinate with the source owner on its OAuth and user-scoped authorization configuration.
Recommended Free Tools
- Identify the precise source, content scope, and intended Microsoft 365 experience.
- Confirm eligible licenses for the users who need the connector’s results.
- Determine whether source permissions can be preserved and mapped correctly for each user.
- Decide what content should be in scope and how current it needs to be.
- Assign an AI administrator and obtain any required source-side administrative access.
- Use the deployment guide for the exact connector; configuration varies by source.
Deploy a synced connector without oversharing
Access permissions are a creation-time decision for a synced connector. Microsoft documents two choices: respect the source access-control lists (ACLs), so only people with access to the source data can see it, or make the content visible to everyone in the organization. The latter can expose sensitive material to users who should not see it.
Rank #3
- Choose the connector and prepare the source. Confirm that the source has a suitable connector, collect required credentials, and review the source’s environment and content permissions.
- Configure it in the Microsoft 365 admin center. An AI administrator deploys the connector; follow that connector’s deployment guide for source-specific settings.
- Select the access option deliberately. Prefer respecting source ACLs when access must follow existing source permissions. Choose organization-wide visibility only when all organizational users are meant to see the indexed content.
- Validate before broad use. Check results with representative users who have different source permissions and confirm that restricted content is not visible to unauthorized users.
Microsoft says connector access permissions cannot currently be updated after creation. If the selected permission model is wrong, delete the connection and recreate it with the intended access setting; do not assume changing source permissions alone corrects a connector-level choice that was made incorrectly.
Use federation when queries should stay at the source
A federated connector does not ingest or index the external content. Copilot queries the source at runtime, so the source must be available to answer those queries and support compatible authentication and query patterns. The implementation must ensure the request is authorized for the signed-in user rather than relying on an overly broad shared identity.
For a custom federated connector, Microsoft’s setup guidance describes an MCP server that exposes read-only tools, such as search, fetch, or query, with OAuth-based identity. Treat the server and its user-scoped authorization as part of the security boundary. Have the source owner review what each tool can return and verify that its authorization reflects the requesting user’s access.
Connect an on-premises SharePoint Server farm
On-premises SharePoint Server uses a dedicated connector route rather than a generic assumption that cloud connectors can reach an internal farm. Microsoft’s deployment guidance calls for a configured SharePoint Server environment, a Microsoft Graph connector agent installed and registered to the tenant, and an AI administrator to deploy the connector. Authentication and permission configuration depend on the environment.
- Confirm the SharePoint Server environment is configured for the connector deployment.
- Install and register the Microsoft Graph connector agent to the Microsoft 365 tenant.
- Prepare the environment-specific authentication and permissions with the SharePoint and identity administrators.
- Deploy the SharePoint Server connector as an AI administrator and test results with users who have different permissions.
Use Microsoft’s SharePoint Server deployment guide for the exact agent and environment steps. The required configuration is specific to the farm and authentication model.
Test the result with real permission differences
Do not validate a connector only with the administrator account that created it. Test with representative users, including users who should have access to the source content and users who should not. For a synced connector, check that expected content appears after indexing and that restricted content stays restricted. For a federated connector, test live queries and confirm that authorization follows each signed-in user.
- Expected content is missing: Check source credentials, source configuration, whether the content is in scope, sync or query behavior, and whether the user is eligible and authorized.
- Results are stale: For a synced connector, account for its recurring schedule rather than expecting every source change to appear immediately. For federation, check source availability and its runtime query behavior.
- Unexpected users can see content: Stop broad rollout and review the connector’s access choice and source ACL mapping. If the connector was created with the wrong permission setting, recreate it with the intended model.
- Federated queries fail: Have the source and identity administrators verify OAuth configuration, user-scoped authorization, and compatibility with the connector’s required query pattern.
Because connector catalogs, licensing, and setup guidance can change, recheck the current Microsoft prerequisites and the source-specific deployment guide when implementing or revisiting a connection.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
