Microsoft Restricted Claude Fable 5 for Employees Due to Data Policy

Anthropic released Claude Fable yesterday, its first Mythos-class model, but problems have already emerged inside Microsoft. The company restricted employee access to Claude Fable 5 due to new Anthropic requirements regarding data storage and processing.

Quick Release for Customers

Microsoft quickly deployed Claude Fable 5 for its customers on GitHub Copilot and Azure Foundry — updates were released the same day. The public release went smoothly, and the model became available in the model selection interface for all users. This shows that Microsoft and Anthropic coordinated well before launch. For GitHub Copilot and Foundry customers, access to Fable represents a meaningful improvement. The model promises better performance compared to previous Claude versions, and Microsoft's enthusiasm was fully justified.

But Inside the Company — a Ban

Inside Microsoft itself, the situation is different. According to sources from The Verge, Claude Fable 5 is absent from the model picker in the internal version of GitHub Copilot used by employees. Microsoft workers cannot select Fable, even though the company just offered it to its customers. This decision was made by Microsoft itself, not Anthropic. All other Claude models — including older versions — remain available for internal use. The company chose not to provide its team with access to the latest model.

Zero Data Retention: The Key Point of Contention

The main reason is data processing policy. All Claude models, except Fable, operate under Zero Data Retention (ZDR) mode. This means Anthropic does not store or use data transmitted through the API for model training or other purposes. The ZDR commitment is a critical point in Microsoft's contract with Anthropic. The company requires this mode for all enterprise customers, including Azure Foundry. Claude Fable comes with different data processing terms that are incompatible with ZDR policy. Microsoft rejected this.

"Employee data requires maximum protection"

Why This is Critical

For Microsoft, this is not simply a technical issue — it is risk management:

• Employee data (source code, documents, technical interactions) is more sensitive than customer data
• Storing corporate information requires the highest level of protection
• ZDR was chosen as the only acceptable mode for internal use
• Deviating from ZDR would create a precedent for other models and vendors

Microsoft may reconsider the terms, but for now the blockade remains.

What This Means

The conflict shows growing tension between the desire to use the latest AI models and the need to protect sensitive corporate data. Microsoft may be the first major company to face this dilemma in the context of Fable, but certainly not the last. The question remains open: will Anthropic reconsider its terms, or will Microsoft agree to a new level of risk?