Microsoft is moving to reduce its reliance on OpenAI by striking a deal to use Anthropic’s artificial intelligence models in its Office 365 applications, according to a report by The Information. The arrangement will see Anthropic’s Claude models power new features in Microsoft Word, Excel, Outlook, and PowerPoint, alongside OpenAI’s technology. This marks a significant shift for Microsoft, which has until now depended solely on OpenAI’s ChatGPT models for its productivity suite.
The decision comes against the backdrop of rising tensions between Microsoft and OpenAI. While Microsoft has invested heavily in OpenAI, the startup has simultaneously pursued its own infrastructure projects and is reportedly working on a LinkedIn competitor, underscoring a growing independence. At the same time, Microsoft is negotiating a new deal with OpenAI as the latter considers restructuring into a for-profit entity. However, sources insist that the Anthropic agreement is not a bargaining tactic but a strategic move to diversify.
Reports suggest Microsoft leaders see advantages in Anthropic’s latest model, Claude Sonnet 4, particularly in producing visually appealing PowerPoint presentations. This is not Microsoft’s first foray into diversifying AI partnerships; through GitHub Copilot, the company already offers models from rivals like Anthropic and xAI. Microsoft has also begun developing its own systems, unveiling MAI-Voice-1 and MAI-1-preview as part of its bid for greater independence in the AI arms race.
Meanwhile, OpenAI is charting its own path to reduce dependence on Microsoft. Recently, it launched a jobs platform designed to rival LinkedIn and is preparing to manufacture AI chips in collaboration with Broadcom by 2026. This would enable OpenAI to operate on its own hardware rather than rely entirely on Microsoft’s Azure cloud. Despite these developments, Microsoft maintains that its partnership with OpenAI remains intact, stressing that both companies will continue to collaborate on advancing frontier AI models