Microsoft has added Anthropic’s Claude artificial intelligence models to its 365 Copilot enterprise platform, enabling organisations to choose between different AI providers for workplace automation and research tasks.
The integration introduces Claude Sonnet 4 and Claude Opus 4.1 models alongside existing OpenAI systems, initially available through Microsoft’s Researcher agent and Copilot Studio development environment. Users can now select preferred AI models for specific business functions rather than being restricted to single-vendor solutions.
Microsoft’s Researcher agent, designed for complex analytical tasks, now supports both OpenAI and Anthropic models. The system processes multistep research by analysing web content, third-party data sources and internal company information, including emails, documents, meetings and chat records.
Copilot Studio gains access to both Claude models for building custom enterprise AI agents. The platform supports mixed deployments where organisations can assign different AI models to specialised tasks, combining Anthropic, OpenAI and Azure Model Catalog options within single workflows.
The service launches through Microsoft’s Frontier Program for existing 365 Copilot licence holders who request access. Organisation administrators must enable the feature through Microsoft’s admin centre before deployment.
Microsoft confirmed that Anthropic models operate outside Microsoft-managed infrastructure and remain governed by Anthropic’s terms of service, indicating a partnership arrangement rather than direct platform integration.
The move addresses enterprise demand for AI model flexibility as organisations evaluate different systems for accuracy, performance and cost-effectiveness across varying use cases. Anthropic’s Claude models have gained adoption for reasoning-intensive tasks and document analysis applications.
Interface changes accompany the integration, with Researcher offering model selection options and Copilot Studio providing dropdown menus for switching between AI providers. This eliminates requirements for separate platforms when accessing different AI capabilities.
The expansion represents Microsoft’s strategy to provide enterprise customers with AI model choice while maintaining its primary partnership with OpenAI. Companies can now optimise AI selection for specific business requirements.
Microsoft indicated plans to expand Anthropic model integration across additional 365 Copilot features, though specific timelines and applications remain undisclosed.
