Microsoft Blocks Employee Access to Anthropic's Claude Fable 5 Over Data Retention Concerns
Summary
Microsoft blocks internal employee access to Anthropic's Claude Fable 5 after the new AI model's 30-day data retention policy — and up to two years for flagged content — raises legal concerns over confidential data, while external GitHub Copilot and Foundry customers retain access as Microsoft's legal team evaluates the changes.
Key Points
- Microsoft is restricting internal employee access to Claude Fable 5, Anthropic's newly released Mythos-class AI model, due to concerns over its new data retention requirements.
- Unlike other Claude models that operate under Zero Data Retention rules, Claude Fable 5 retains user prompts and outputs for up to 30 days — and up to two years if flagged for policy violations — raising legal concerns around customer and confidential data.
- Microsoft's legal teams are currently evaluating Anthropic's data retention changes before deciding whether to clear Claude Fable 5 for internal use, while the model remains available to external GitHub Copilot and Foundry customers.