Microsoft blocks employee use of Anthropic Claude Fable 5

Microsoft has temporarily blocked its employees from using Anthropic's latest Claude Fable 5 artificial intelligence model internally, citing concerns over the AI startup's updated data retention policy. The restriction was implemented on June 10, 2026, just one day after the model's public launch, while Microsoft continues to offer the same technology to its customers through GitHub Copilot and Azure AI Foundry services.

The company's legal and compliance teams are currently reviewing the implications of Anthropic's new policy for Mythos-class models, which includes Claude Fable 5. The core issue is a mandatory 30-day storage window for all user prompts and generated responses. Anthropic states this data is automatically deleted after 30 days unless flagged for safety investigations or legal obligations, in which case retention can extend up to two years. The company emphasizes the stored information is not used to train future AI models.

This policy represents a significant shift from the zero-retention options typically offered with enterprise AI tools, where conversations are deleted immediately after use. For Microsoft, the concern centers on the potential exposure of...