Microsoft embeds Anthropic in Office

- Microsoft has integrated Anthropic’s Claude models into Microsoft 365 Copilot and related Office workflows, broadening model choice beyond OpenAI as of April and May 2026. - Microsoft said Anthropic models are available in Microsoft 365 Copilot, Researcher, Copilot Studio and Excel agent mode, though disabled by default in the UK and EU. - Anthropic said Outlook support is in public beta, while Word, Excel and PowerPoint add-ins became generally available on May 7.

Microsoft has added Anthropic’s Claude models to parts of its Microsoft 365 and Office software stack, expanding the set of AI models available inside products long associated with OpenAI. Microsoft documentation published in April said Anthropic models would be available by default in certain regions across Microsoft 365 Copilot, Researcher, Copilot Studio, Power Platform, agent mode in Excel, and Word, Excel and PowerPoint agents. Anthropic separately said on May 7 that Claude add-ins for Excel, PowerPoint and Word were generally available, with Outlook in public beta. The moves place Anthropic software inside core workplace applications used for documents, spreadsheets, presentations and email. ### Where exactly is Claude showing up inside Microsoft’s software? Microsoft’s support documentation says Anthropic models can be used in Microsoft 365 Copilot on web, desktop and mobile, and in Copilot Studio, Researcher, Power Platform and agent mode for Excel. The same document says Word, Excel and PowerPoint agents can also use Anthropic models, with interface indicators showing when Claude is in use. (learn.microsoft.com) Anthropic’s own product pages describe a parallel set of integrations aimed at end users. Anthropic said Claude works inside Excel, PowerPoint, Word and Outlook, and that context can carry across those four apps as users move from email to documents and slides. Anthropic said the Word, Excel and PowerPoint add-ins became generally available on May 7, while Outlook entered public beta for paid plans the same day. ### Is this a Microsoft product decision, an Anthropic add-in, or both? (learn.microsoft.com) April 16 is the clearest date for Microsoft’s own model-expansion push. In a Microsoft 365 Copilot blog post, Bas Brekelmans wrote that Anthropic’s Claude Opus 4.7 was being added to Copilot Cowork (Frontier) and Copilot Studio early release environments and was rolling out to Copilot in Excel. (claude.com) May 7 is the clearest date for Anthropic’s direct Office integrations. Anthropic said its Claude add-ins for Excel, PowerPoint and Word were generally available, and Outlook support was opening in beta, giving users a way to work with Claude across Microsoft 365 applications with persistent context between them. ### What has Microsoft said about using more than one model? Microsoft’s documentation says the Anthropic update “gives users in your organization the ability to use multiple AI models” across Microsoft offerings. (techcommunity.microsoft.com) The company said the change reflects a commitment to offering “choice between leading AI models” while keeping Microsoft’s enterprise security and compliance controls in place. (claude.com) Benzinga reported on May 21 that Microsoft had already integrated Claude into Office 365 applications including Word, Excel, Outlook and PowerPoint, describing the move as part of a broader AI strategy beyond the company’s longstanding relationship with OpenAI. That report also said Anthropic had been increasing its Azure server rentals. Reuters did not independently confirm the Azure detail in the material reviewed here. (learn.microsoft.com) ### Are there limits on where customers can use these Anthropic models? Microsoft’s April documentation says Anthropic models in Microsoft offerings are excluded from the EU Data Boundary and, where applicable, in-country processing commitments. The company said customers within the EU Data Boundary and customers in the UK will have Anthropic models disabled by default. That regional restriction is one of the more concrete signs that Microsoft is treating Anthropic as one model provider among several inside a controlled enterprise framework, rather than as a universal default. (benzinga.com) Microsoft’s documentation ties the rollout to tenant-level controls and product-specific indicators rather than a single companywide switch. ### Why is this landing as OpenAI’s legal activity expands? (learn.microsoft.com) Reuters reported on May 21 that OpenAI had expanded its roster of law firms for high-stakes lawsuits and deals, including defense against a lawsuit by xAI alleging that OpenAI and Apple monopolize markets for smartphones and generative-AI chatbots. Reuters said Wachtell and Morrison & Foerster helped OpenAI and Chief Executive Sam Altman defeat Elon Musk’s lawsuit earlier that week. (learn.microsoft.com) Microsoft and Anthropic have not tied the Office integrations to OpenAI’s legal posture. But the timing places Anthropic inside Microsoft’s productivity products as OpenAI manages a widening set of legal and commercial disputes around distribution, partnerships and corporate structure, according to Reuters. (economictimes.indiatimes.com) ### What comes next for users and administrators? Microsoft’s current documentation says Anthropic models are already enabled by default in certain regions and that administrators can expect user-interface indicators when Claude is selected in Microsoft 365 Copilot. Anthropic’s next visible milestone is Outlook: the company said on May 7 that Outlook support was in public beta for paid plans, while Word, Excel and PowerPoint were already generally available. (learn.microsoft.com) (economictimes.indiatimes.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.