Information reports reveal that Microsoft will use the Anthropic AI model in Office 365 applications. Anthropic will provide new functionality in Word, Excel, Outlook and PowerPoint applications, which means that Anthropic’s AI will take over Microsoft’s market from its competitor OpenAI and may also mark the end of Microsoft’s earlier reliance on OpenAI.
In a growing rift with Openai, Microsoft is promoting the diversification of its AI partnership, which is pursuing its own infrastructure projects and potential LinkedIn competitors.
When Microsoft negotiated a new agreement with OpenAI to ensure a profitable reorganization of access, a new deal was made with OpenAI to ensure access to its AI model. But the message is that this is not a negotiation strategy. Microsoft’s leaders believe that the latest version of Anthropic (Claude Sonnet 4), especially in certain functions, perform better than OpenAI, for example, by creating a beautiful aesthetic PowerPoint presentation.
However, this is not the first Microsoft branch. Although OpenAi is a default model, Microsoft provides other models through Github Copilot, such as Grok in XAI and Claude in Anthropic. Microsoft also tried to promote self-study AI models, which recently launched two internal models under the flag: Mai-Voice-1 and Mai-1-preview.
At the same time, Openai also tried to get out under Microsoft’s influence. Last week, Openai launched an employment platform to take over from Microsoft’s LinkedIn, and the Financial Times reported that OpenAI, in collaboration with Broadcom, would begin large-scale production on its first AI chip in 2026.
Microsoft spokesman Michael Collins told TechCrunch: “As we say, Openai will continue to be a partner in our Frontier model, and we remain committed to long-term partnerships.”
