Microsoft adds Anthropic AI to Copilot 365 — after claiming OpenAI's GPT-4 model is "too slow and expensive"

There's no doubt that Microsoft's multibillion-dollar partnership with OpenAI has been under immense strain and tension, especially with the latter's plan to evolve into a for-profit venture to avoid losing investor funding, outsider interference, and hostile takeovers.

Over the past few months, Microsoft has made strategic moves, seemingly relieving itself from an overdependence on OpenAI and its tech by developing its own off-frontier models, which will be 3 to 6 months behind the ChatGPT maker's.

More recently, Microsoft's AI CEO, Mustafa Suleyman, highlighted the company's plans to make "significant investments" in building and developing its own AI chip cluster to unlock self-sufficiency in AI. While this play is still in the works, Microsoft just announced its plan to integrate Anthropic's AI models into Copilot.

This doesn't mean that Microsoft is severing its ties with OpenAI. The ChatGPT maker's latest AI models will continue to be available to users through Copilot. However, you can switch between OpenAI and Anthropic's AI models depending on your preference and needs.

It's worth noting that Anthropic's models are shipping to Microsoft 365 Copilot's Researcher agent to help generate reports and analyze data. If you're in Microsoft's Frontier program, which allows users to gain early access to AI features, you should be able to access Anthropic's Claude Sonnet 4 and Claude Opus 4.1 directly from Copilot, aside from OpenAI's latest offerings.

According to Microsoft’s president of business and industry Copilot, Charles Lamanna:

“And stay tuned: Anthropic models will bring even more powerful experiences to Microsoft 365 Copilot.”

OpenAI didn't cut it for Microsoft

This week, NVIDIA announced its plans to invest up to $100 billion in OpenAI to facilitate the construction of massive new data centers, which will help meet its high demand for compute.

As you may know, Microsoft was OpenAI's exclusive cloud provider partner, but it lost the status earlier this year after the AI firm unveiled its $500 billion Stargate project designed to facilitate the construction of data centers across the United States. However, the tech giant still holds the right of first refusal since it's OpenAI's largest backer and investor.

There have been signs indicating that Microsoft is looking beyond OpenAI, with multiple reports highlighting its plans to integrate third-party models in Copilot. Interesting, Microsoft indicated that OpenAI's GPT-4 model is "too slow and expensive" to meet Copilot 365 users' needs.

Similarly, OpenAI has been looking beyond Microsoft for its cloud compute needs, with reports suggesting that it is in discussions with Google for additional computing support. This is despite Sam Altman claiming that OpenAI is no longer compute-constrained after Microsoft wiggled out two mega data center deals because it didn't want to provide additional training support for ChatGPT.

Salesforce CEO Marc Benioff's prediction is seemingly starting to take shape. After OpenAI unveiled its Stargate project, the executive indicated that Microsoft won't use OpenAI's tech in the future.

Follow Windows Central on Google News to keep our latest news, insights, and features at the top of your feeds!