January 23, 2026 Microsoft is quietly broadening its internal use of Anthropic’s Claude Code, encouraging not just software engineers but also designers, product managers and other nontechnical employees to use the AI coding tool across some of the company’s most important teams. According to people familiar with the rollout, Microsoft has begun asking thousands of employees to experiment with Claude Code, even if they do not write code for a living.
Microsoft’s relationship with Anthropic is not new. The company began adopting Anthropic’s Claude Sonnet 4 model within its developer division in mid-2025, later favoring it for certain paid GitHub Copilot users. What’s changed now is the scale and scope of adoption. Claude Code is being tested across major engineering organizations, including Microsoft’s CoreAI team, led by former Meta engineering executive Jay Parikh, and the Experiences + Devices division responsible for Windows, Microsoft 365, Outlook, Teams, Edge, Bing and Surface.
Employees without formal programming backgrounds are being encouraged to use Claude Code for prototyping and experimentation, allowing ideas to move faster from concept to working software. Microsoft has also approved the tool for use across all internal codebases within its Business and Industry Copilot teams.
Internally, Microsoft engineers are being asked to use both Claude Code and GitHub Copilot in parallel and provide structured feedback comparing the two tools. While Microsoft continues to sell GitHub Copilot as its primary AI coding assistant to customers, the breadth of the Claude Code pilots suggests the company is actively evaluating alternatives and potentially future offerings.
Microsoft has become one of Anthropic’s largest customers, according to recent reporting. It also now counts Anthropic model usage toward Azure sales quotas, an unusual move for a company that typically prioritizes its own products or those built by long-time partner OpenAI. In November, Microsoft and Anthropic signed a deal that allows Azure customers to access Claude Sonnet 4.5, Claude Opus 4.1, and Claude Haiku 4.5, alongside Anthropic’s commitment to purchase $30 billion worth of Azure compute.
Microsoft has also begun favoring Anthropic models inside certain Microsoft 365 and Copilot features where they have outperformed OpenAI’s equivalents. Still, Microsoft says it views this experimentation as part of a broader, multi-model strategy.
As Frank Shaw, Microsoft’s communications chief, said in a statement, “Companies regularly test and trial competing products to gain a better understanding of the market landscape. OpenAI continues to be our primary partner and model provider on frontier models, and we remain committed to our long-term partnership.”
