Microsoft has announced the integration of Anthropic’s AI models into its Copilot assistant, providing business users with more choices beyond OpenAI’s offerings. Starting September 24, 2025, enterprise customers can opt to use Anthropic’s Claude Sonnet 4 and Claude Opus 4.1 in specific Copilot features, including the Researcher agent and when building custom agents in Copilot Studio. For AI developers, enterprise IT leaders, and tech observers searching Microsoft Anthropic Copilot integration, Claude models in Copilot Studio, or Microsoft AI diversification 2025, this update reflects Microsoft’s strategy to reduce dependence on a single provider while enhancing Copilot’s reasoning and workflow capabilities. Administrators must enable the models, and they run on Amazon and Google clouds under Anthropic’s terms.
The rollout begins with Microsoft’s Frontier program, with broader preview access in two weeks and production readiness by year-end.
Integration Details: Claude in Researcher and Copilot Studio
Copilot remains powered by OpenAI’s latest models by default, but users can now switch to Anthropic’s Claude for tasks requiring advanced reasoning and orchestration. Claude Opus 4.1 is particularly suited for complex research, while Sonnet 4 excels in workflow automation.
- Researcher Agent: Users in the Frontier program see a “Try Claude” button to toggle models for building reports and gathering information.
- Copilot Studio: A dropdown menu allows selection of Claude models for creating enterprise-grade agents, with automatic fallback to OpenAI if disabled.
- Admin Controls: Enabled via the Microsoft 365 admin center; once activated, users can mix models for tasks like orchestration and tool use.
Charles Lamanna, President of Microsoft’s Business and Industry Copilot, wrote: “Anthropic models will bring even more powerful experiences to Microsoft 365 Copilot.”
| Feature | OpenAI Default | Anthropic Option | Availability |
|---|---|---|---|
| Researcher Agent | GPT-4o | Claude Opus 4.1 | Frontier Program (Today) |
| Copilot Studio | GPT-4o | Claude Sonnet 4 / Opus 4.1 | Frontier (Today); Preview (2 Weeks) |
Strategic Shift: Diversifying Beyond OpenAI
Microsoft’s integration of Anthropic models signals a broader diversification strategy, following last year’s addition of Anthropic and Google models to GitHub Copilot Chat. Despite a $13 billion investment in OpenAI, Microsoft is hedging with rivals like Anthropic (valued at $183 billion) and in-house models like MAI-1-preview.
- Anthropic’s Edge: Claude models are hosted on AWS and Google Cloud, subject to Anthropic’s terms rather than Microsoft’s.
- Broader Ecosystem: Microsoft now offers models from xAI and Meta in its data centers, emphasizing flexibility.
- Enterprise Focus: Aimed at Microsoft 365 Copilot users, with automatic fallback to OpenAI if Anthropic is disabled.
Implications: Enhanced Flexibility and Competition
This addition bolsters Copilot’s versatility for tasks like research and agent building, where Claude’s reasoning strengths complement OpenAI’s. It signals Microsoft’s evolution from OpenAI exclusivity, potentially pressuring rivals to innovate faster.
- User Benefits: Choice in models for optimized workflows, with no additional configuration for fallbacks.
- Market Dynamics: Anthropic’s $183 billion valuation rises with Microsoft tie-ins; OpenAI remains default but faces competition.
- Future Rollout: More Anthropic experiences in Copilot, plus in-house MAI-1-preview testing.
Conclusion: Microsoft’s Multi-Model Copilot Evolution
Microsoft’s addition of Anthropic AI to Copilot is a strategic diversification, empowering enterprises with Claude options alongside OpenAI for superior reasoning and agent building. As rollout expands, it heralds a more flexible AI future. For those leveraging Copilot, enable it now—will Claude outshine GPT in workflows? The toggles await. CNBC
