Microsoft is integrating an automatic AI model selection feature into Visual Studio Code that will choose the best-performing model for specific coding tasks. This new functionality will dynamically assign models such as Claude Sonnet 4, GPT-5, GPT-5 mini, and others to GitHub Copilot users based on performance. While free users will have access to a range of models, paid subscribers will primarily rely on Claude Sonnet 4. This shift signals a strategic preference by Microsoft for Anthropic’s AI models over OpenAI’s latest offerings, including GPT-5. Internal sources familiar with Microsoft’s developer strategy confirm that the company has been directing its own engineering teams to use Claude Sonnet 4 in recent months. In a June internal email, Julia Liuson, head of Microsoft’s developer division, stated, “Based on internal benchmarks, Claude Sonnet 4 is our recommended model for GitHub Copilot.” This guidance was issued before the GPT-5 release, and Microsoft has not updated its recommendation since. The move underscores a growing divergence in how Microsoft is deploying AI across its products. Beyond GitHub Copilot, Microsoft is reportedly planning to incorporate Anthropic’s models into select features of Microsoft 365 Copilot. According to The Information, internal testing showed that certain Anthropic models outperformed OpenAI’s in applications like Excel and PowerPoint, prompting the shift. Microsoft is also investing heavily in its own AI infrastructure. In a recent employee town hall, AI chief Mustafa Suleyman revealed that the company is building a large-scale AI training cluster. He noted that the current MAI-1-preview model was trained on just 15,000 H100 GPUs—a relatively small setup compared to the scale of future plans. Despite these developments, Microsoft and OpenAI continue to maintain a close partnership. A new agreement announced last week may pave the way for OpenAI’s eventual public offering. Since 2019, Microsoft has invested over $13 billion in OpenAI and operates under a complex revenue-sharing arrangement. The updated deal allows OpenAI to use rival cloud providers, marking a shift in their long-standing exclusivity. Microsoft is expected to share more details about the next phase of its relationship with OpenAI in the coming weeks. However, the increasing reliance on Anthropic models across key developer and productivity tools reflects a broader strategic pivot—balancing deep collaboration with OpenAI while selectively adopting superior alternatives from other AI labs.