Microsoft 365 Copilot now sends data to Anthropic by default. Here's how to think about it.

If you use Microsoft 365 Copilot, you gain a new data subprocessor on 7 January 2026: Anthropic. A subprocessor is a second company that handles your data on Microsoft’s behalf. Microsoft has enabled Claude models by default across many Copilot experiences, under the existing Product Terms and Data Protection Addendum (DPA). The official Microsoft Learn page is here: Connect Microsoft 365 Copilot to AI subprocessors.

The opt-out is the bit that matters most. If you do nothing, the integration is on. When Copilot answers a request, your prompt can now be sent to Anthropic, and so can the Microsoft 365 data that prompt pulled in.

For most businesses this is a governance event. Worth at least one structured conversation between your IT, your data-protection lead, and whoever signs off on cross-border data transfers.

Your data passes through a second company

Anthropic is a separate company from Microsoft. When Claude handles a Copilot prompt, the data needed to answer it leaves Microsoft’s systems and goes to Anthropic’s. That changes who you are trusting with the data, in three ways:

• Anthropic joins your list of data processors. Any record you keep of who processes your data needs to name them.
• It can be a cross-border transfer. Anthropic’s processing locations and Microsoft’s aren’t the same. You may be subject to data-sovereignty rules, or to the Australian Privacy Principles on overseas disclosure. If so, treat this as the cross-border transfer it is, and not as another Microsoft feature running inside your own systems.
• Your existing DLP and Purview rules may not have been written with this subprocessor in mind. The classification, redaction and policy work that kept regulated content out of Microsoft’s external models now needs reviewing against Anthropic’s processing too.

None of this says Anthropic is the wrong model provider. Claude is a strong model. But “default on” doesn’t mean the legal and governance work is already done.

The trade-off Microsoft is offering

Disabling Anthropic isn’t free. Several Copilot features use Claude specifically:

• The Researcher agent.
• Custom agents built in Copilot Studio.
• Office “Frontier” experiences.

Turn Anthropic off and you get less from the Copilot licences you’ve already paid for. That’s the real trade: the newest capability now, or governance-controlled capability later.

For most regulated businesses, disable it by default. That means legal, finance, health, and anyone holding personal information at scale. Run the legal and data-governance review, then re-enable selectively where the business case supports the extra processor.

If your business isn’t regulated and your data is low-sensitivity, leaving it on and using the features might be the better call. Make sure that’s an actual decision and not a default.

What we’re advising clients

A short checklist for the next two weeks:

• Inventory your Copilot use. Who has Copilot licences, what they use them for, and what data those workflows touch.
• Turn Anthropic off in the Microsoft 365 admin centre while you work through the rest of this list. The setting is reversible.
• Review your DPA register. Add Anthropic. If you don’t have a register yet, this is the moment to start one.
• Run a Purview and DLP review against the workflows that would route through Copilot. Make sure your sensitivity labels actually catch the content you’d want held back from any external model.
• Make a written decision about which user groups can use Anthropic-backed Copilot features and which can’t. Document the reasoning and the data classes you’re protecting.
• Re-enable selectively only where the workflow genuinely benefits and the data class is appropriate.

If you’re a managed compliance client, your account lead will raise this at the next review anyway. Want a hand thinking it through before then? Get in touch.