Microsoft Shifts Gears: Favors Anthropic's AI Over OpenAI in Key Products

Microsoft's Evolving AI Strategy

Microsoft is reportedly shifting its AI strategy, now prioritizing Anthropic's Claude Sonnet 4 model over OpenAI's GPT-5 for key products, including its flagship developer tool, Visual Studio Code. This marks a notable change despite Microsoft's substantial investments in OpenAI 1

1

.

Claude Sonnet 4's Performance Edge

In Visual Studio Code, an automatic AI model selection feature will be introduced for GitHub Copilot, with Claude Sonnet 4 recommended for paid users based on internal benchmarks. Julia Liuson, head of Microsoft's developer division, stated this preference in an internal email, highlighting Claude Sonnet 4's superior performance for coding assistance even after the release of OpenAI's GPT-5 models 2

2

. Microsoft also plans to integrate Anthropic's models into Microsoft 365 apps like Excel and PowerPoint, where they have reportedly outperformed OpenAI alternatives in productivity scenarios 1

1

.

Diversifying AI Investments

Beyond external partnerships, Microsoft is bolstering its internal AI capabilities. Mustafa Suleyman, Microsoft's AI chief, confirmed significant investments in the company's own AI infrastructure. He noted that their MAI-1-preview model was trained on a cluster of 15,000 H100 GPUs, hinting at future plans for even larger scale AI development 3

3

.

Partnership Dynamics with OpenAI

This strategic pivot raises questions regarding Microsoft's long-term relationship with OpenAI, an entity it has invested over $13 billion in since 2019. Despite the close ties, Microsoft's pragmatic approach to selecting the best-performing AI models for specific applications suggests a diversified strategy. Recently, Microsoft and OpenAI announced a new agreement, potentially facilitating OpenAI's initial public offering and allowing OpenAI to utilize rival cloud providers, with further details expected soon from Microsoft 1

1

.