Summary
- Anthropic models will become the default for many customers in Microsoft 365 Copilot, subject to new data processing and product terms.
- Admins will receive a new toggle in the Microsoft 365 admin center, with most commercial customers enabled by default.
- Exclusions apply for certain regions like the EU/EFTA and UK, where customers must opt in to use Anthropic models.
- Admins currently have the option to opt out of using Anthropic models before the default change occurs.
- Specific features may be restricted if Anthropic models are disabled.
Admin Impact: High
User Impact: High
Release Start: 07 Jan 2026
Release End: 07 Jan 2026
Services: M365 Copilot
Category: Plan for change
Tags: Feature Update, User Adoption, Admin Action
History
12/8/2025 Item Added to Message Center
Microsoft Message
Introduction
In September, Anthropic models became available in certain Copilot experiences through an opt-in setting, where use was subject to Anthropic’s commercial terms and data processing agreement. On January 7, 2026, Anthropic models will be enabled by default for many customers and now offered under the Microsoft Product Terms and Data Protection Addendum. To enable this change, Anthropic has been onboarded as a Microsoft subprocessor. This transition extends Microsoft’s contractual commitments and data protection standards to the use of Anthropic models in Copilot experiences.
How this affects your organization
Who is affected:
- Admins managing Microsoft 365 Copilot and Microsoft Copilot Studio.
- Microsoft 365 Copilot licensed customers participating in select Frontier features such as Word, Excel, PowerPoint, and Researcher agents, as well as Agent Mode in Excel.
Key changes
- Dec 8, 2025: New admin toggle for Anthropic models appears in the Microsoft 365 admin center (MAC). For most commerical cloud customers, it will be enabled ON by default (except customers in EU/EFTA and UK will be OFF by default). See “customer and region exclusions” section below for more details.
- January 7, 2026: Anthropic as a Microsoft subprocessor becomes enabled on your tenant; the legacy admin toggle to opt-in to Anthropic’s commercial terms and data processing agreement is deprecated.
- If you previously opted in to the legacy Anthropic toggle, and you’re in an excluded region, you would need to opt in to the new subprocessor toggle.
- Admins can opt out in MAC now – no need to wait until January.
Some features are only available when Anthropic models are enabled. If you turn off this option, certain features may no longer be accessible.
Customer and region exclusions:
The following customers and regions will not have the toggle set to default on:
- Customer tenants in EU/EFTA and UK: The toggle will be set to default off. You will need to opt in to use Anthropic models as a subprocessor.
- Government clouds (GCC, GCC High, DoD) or sovereign clouds: Anthropic models are not yet available in government or sovereign clouds. No toggle will be present for those clouds.
What does it mean that Anthropic is now a Microsoft subprocessor?
As a subprocessor for Copilot experiences, Anthropic will operate under Microsoft’s direction and contractual safeguards. This includes coverage under the Microsoft Data Protection Addendum (DPA) and Product Terms. In addition, use of Anthropic models in Microsoft 365 Copilot falls under our Enterprise Data Protection as described here. Note that Anthropic models are currently excluded from EU Data Boundary and when applicable, in-country processing commitments.
Learn more:
- Integrate and manage Anthropic’s AI models in Microsoft 365 Copilot
- Enterprise data protection in Microsoft 365 Copilot and Microsoft 365 Copilot Chat
- Use Claude with Researcher in Microsoft 365 Copilot
- Get started with Word, Excel, and PowerPoint Agents (Frontier) in Microsoft 365 Copilot
- Choose an external model as the primary AI model – Microsoft Copilot Studio