Microsoft Reportedly Restricts Claude Fable 5 Use Over Anthropic Data Retention Policy

Microsoft is reportedly limiting employee use of Claude Fable 5 because of concerns over Anthropic’s updated data retention requirements, according to The Verge.

The temporary restriction comes as Microsoft’s legal and compliance teams review how Anthropic stores prompts, outputs, and content flagged by its safety systems.

According to the report, Anthropic’s new requirements allow prompts and outputs to be retained for up to 30 days. Content flagged by its safety systems can reportedly be kept for up to two years.

Microsoft worries about sensitive data exposure

Microsoft is concerned that employees could accidentally share confidential information while using Claude Fable 5.

That data could include customer information, internal corporate material, business plans, code, or other sensitive company details.

The concern becomes more serious if Anthropic’s safety systems flag that content. In that case, the data could reportedly become available during investigations or enforcement reviews.

Microsoft also sees a competitive risk because Anthropic operates as a rival AI company.

Claude Fable 5 restriction appears temporary

The restriction appears to be temporary while Microsoft reviews the compliance impact of Anthropic’s updated retention policy.

The company has reportedly asked employees not to use Claude Fable 5 during this review period.

The move does not necessarily mean Microsoft has made a final decision on Anthropic’s tools. Instead, it suggests the company wants to assess legal, compliance, and data security risks before allowing broader use.

Microsoft is also ending Claude Code licenses

Separately, Microsoft later announced that all Claude Code licenses would end on June 30.

That decision was reportedly not directly tied to the Claude Fable 5 data retention issue.

Reports suggest Microsoft wants employees to move from Claude Code to GitHub Copilot CLI instead. This would keep more developer activity inside Microsoft’s own AI and coding ecosystem.

There may also be a financial angle. Microsoft’s fiscal year ends on June 30, which could make the license termination part of broader cost control or budget planning.

Compliance, competition, and cost may all play a role

Microsoft’s reported Claude Fable 5 restriction appears to involve several factors at once.

The biggest issue is data retention. If employees enter sensitive company or customer information into a third-party AI tool, Microsoft needs to understand where that data goes and how long it remains stored.

The competitive angle also matters. Anthropic competes with Microsoft-backed AI products, including tools connected to GitHub Copilot.

At the same time, ending Claude Code licenses could help Microsoft reduce overlapping software costs and push employees toward its own AI developer tools.

For now, the move looks less like a simple product ban and more like a broader review of AI tool usage inside Microsoft.

Via Windows Central