Microsoft has taken another significant step in its AI strategy by integrating Anthropic’s Claude Sonnet 4 and Claude Opus 4.1 into Microsoft 365 Copilot.
The move, announced this week, broadens the range of AI models available to enterprise customers and enhances the flexibility of Microsoft’s flagship productivity suite.
“Copilot will continue to be powered by OpenAI’s latest models, and now our customers will have the flexibility to use Anthropic models too – starting in Researcher or when building agents in Microsoft Copilot Studio,” said Charles Lamanna, President of Microsoft’s Business and Industry Copilot division.
The addition of Claude Sonnet 4 and Claude Opus 4.1 advances our commitment to bring the best AI innovation from across the industry to Microsoft 365 Copilot”
Until now, Microsoft 365 Copilot experiences were powered exclusively by OpenAI’s GPT models. By introducing Anthropic’s Claude family, Microsoft is positioning its Copilot platform as a multi-model ecosystem, giving organisations the ability to choose the right AI for the right task.
Claude in Researcher: A New AI Option for Knowledge Work
One of the first touchpoints for Anthropic integration is Microsoft’s Researcher agent, designed to support deep research tasks within Microsoft 365.
Researcher users will now see a new “Try Claude” button at the top of their Copilot app, which allows them to opt into using Anthropic’s Opus 4.1 model instead of OpenAI’s.
The experience is designed for seamless switching. Once opted in, users can toggle between OpenAI and Anthropic models depending on the complexity of the task.
Opus 4.1, in particular, is tailored for deep reasoning, making it well-suited for information-heavy workflows where accuracy and context matter most.
Copilot Studio: Building Agents with Anthropic Models
Beyond Researcher, Anthropic’s models are also being introduced into Microsoft Copilot Studio, the low-code platform for creating and managing AI-powered agents.
Copilot Studio users can now build, orchestrate, and deploy agents powered by Claude Sonnet 4 and Opus 4.1. These models enable advanced use cases such as workflow automation, data synthesis, and agent-driven task management.
Crucially, Copilot Studio allows organisations to mix and match models from different providers. This means businesses can design solutions that draw on Anthropic, OpenAI, or even other models from Azure’s model catalog – selecting the best model for each workflow.
The rollout of Claude integration in both Researcher and Copilot Studio is starting today through Microsoft’s Frontier program, available to licensed Microsoft 365 Copilot customers who opt in.
Strategic Implications for Microsoft’s AI Ecosystem
While the addition of Anthropic models strengthens Microsoft 365 Copilot, it also highlights an interesting dynamic in the cloud market. Anthropic’s AI models are currently hosted on Amazon Web Services (AWS) – one of Microsoft’s biggest competitors in cloud infrastructure.
For now, Microsoft accesses Claude through Anthropic’s API, the same way independent developers do. However, given Microsoft’s earlier agreement to host xAI’s Grok 3 models on Azure, industry watchers believe it is only a matter of time before a similar arrangement is struck with Anthropic.
This strategy aligns with Microsoft’s broader push to position Azure as a hub for multi-model AI, where customers can access innovations from across the industry while staying within Microsoft’s cloud ecosystem.
Building on Momentum: Anthropic in GitHub and Beyond
This announcement comes just a week after Microsoft began prioritising Anthropic’s models in another key developer tool: Visual Studio Code.
GitHub Copilot paid users now “primarily rely on Claude Sonnet 4” when using the editor’s new automatic AI model selection feature.
The early performance of Anthropic’s models appears to be convincing Microsoft to expand their footprint further. Reports suggest that Claude models are already being tested in Excel and PowerPoint, with results showing stronger performance in certain tasks compared to OpenAI’s GPT models.
If confirmed, this could represent a major milestone for Anthropic, with its AI embedded in two of the most widely used productivity applications in the world.
What’s Next for Microsoft 365 Copilot?
For Microsoft, the introduction of Anthropic models is not a replacement for OpenAI but an expansion of choice. By giving customers the ability to select from multiple model providers, Microsoft is addressing the growing demand for AI systems that are tailored to specific workflows, industries, and compliance needs.
Lamanna hinted at further developments ahead: “This is just the beginning — we’re committed to delivering model innovation at speed. Anthropic models will bring even more powerful experiences to Microsoft 365 Copilot.”
As AI competition heats up, Microsoft’s strategy is clear: empower customers with flexibility, maintain strong ties with multiple AI providers, and ensure Microsoft 365 Copilot remains at the centre of the enterprise productivity landscape.
