Claude Code spreads fast inside Microsoft teams

Source: theverge.com

TL;DR

The story at a glance

Microsoft has expanded use of Anthropic's Claude Code AI coding tool to big engineering groups, including the division behind Windows and Microsoft 365. Sources inside the company say teams are now told to install it alongside GitHub Copilot for feedback. The push follows months of favoring Anthropic models over OpenAI's in some tools and comes after a November deal giving Foundry customers access to Claude models.[[1]](https://www.theverge.com/tech/865689/microsoft-claude-code-anthropic-partnership-notepad)

Key points

Details and context

Microsoft sells GitHub Copilot externally but runs these pilots to stay sharp on competitors like Anthropic and Cursor. Developers compare tools on real projects across over 100,000 repositories. The company still calls OpenAI its main partner but picks the best model per task, such as Claude in some Microsoft 365 Copilot features.[[1]](https://www.theverge.com/tech/865689/microsoft-claude-code-anthropic-partnership-notepad)

Anthropic's Claude Code and related Cowork tools focus on agent-like help for coding and other work, making them easier for beginners than Copilot in tests. This fits Microsoft's multi-model strategy, unlike its heavy OpenAI tie-up. Junior developers face pressure as AI handles more code creation with less oversight.

Rishi Sunak, now advising both Microsoft and Anthropic, recently discussed AI job shifts with CEO Satya Nadella.

Key quotes

Why it matters

Microsoft's internal shift signals trust in Anthropic's tools challenging its own AI products and OpenAI's lead in coding aids. Engineers and teams get better options now, while customers might see Claude via Azure if pilots succeed; it also hints at job changes from AI agents. Watch for Copilot updates or sales of rival tools, though Microsoft stresses these are just tests.