Microsoft is Expanding Claude Code Usage Among Its Employees

A Different AI Strategy Inside vs. Outside

It has emerged that tech giant Microsoft is in an interesting dilemma regarding AI-powered coding tools. While the company offers GitHub Copilot as its official solution to customers, it is actively encouraging the use of Anthropic's Claude Code within its own organization, particularly among important engineering teams.

Claude Code Pilot Implementation in Core Teams

According to information obtained, Microsoft began adopting the Claude Sonnet 4 model in its developer division last June. The process continued with this model being preferred for GitHub Copilot's paid users. Now, the company has started using Claude Code extensively in its largest engineering teams.

The new AI engineering group CoreAI, led by former Meta engineering chief Jay Parikh, tested Claude Code in recent months. Last week, employees in the Experiences + Devices division, responsible for critical products like Windows, Microsoft 365, Outlook, Teams, Bing, Edge, and Surface, were asked to install Claude Code. Even designers and project managers without coding experience are being encouraged to try this tool to prototype their ideas.

Comparative Use and Future Scenarios

Microsoft software engineers are now expected to use both Claude Code and GitHub Copilot and provide feedback by comparing the two. The company sells GitHub Copilot as its preferred AI coding tool to customers. However, if these large-scale internal pilot programs are successful, there is also a possibility that Microsoft could sell Claude Code directly to cloud customers.

It is stated that Microsoft is now one of Anthropic's largest customers. The company counts sales of Anthropic AI models towards Azure sales quotas. This move is considered unusual given that Microsoft is generally thought to provide incentives to its sales teams only for its own products or OpenAI models.

Partnership Balances and Market Dynamics

Microsoft Chief Communications Officer Frank Shaw, in a statement on the matter, emphasized that companies regularly test competing products to better understand the market. Shaw said, "OpenAI remains our primary partner and model provider for frontier models, and we are committed to our long-term partnership."

While maintaining its commitment to OpenAI, Microsoft is also increasingly working to integrate Anthropic's models and tools both into its own teams and the software it sells to customers. With the agreement signed in November, the two companies provide Microsoft Foundry customers with access to Claude models, while Anthropic committed to purchasing $30 billion worth of Azure computing capacity.

Impacts on Coding and Workforce

The increased use of Claude Code at Microsoft raises the question of what it means for the company's more than 100,000 code repositories. Microsoft stated last year that 91% of its engineering teams used GitHub Copilot and that various teams utilized this AI tool to speed up routine tasks.

While the company's use of AI tools has been largely limited to software engineers, Claude Code and similar tools aim to make AI agent capabilities, coding, and non-coding tasks more accessible. By leveraging Claude Code's ease of use, Microsoft is enabling more non-technical employees to write code using AI. This broad pilot implementation seems poised to reveal the challenges and benefits of this change.

This development also increases pressure on junior developer roles, which are a growing concern in the industry due to AI. Microsoft has taken another step towards a future where control is increasingly taken by more autonomous AI agents from software engineers.