- Mythos-based Claude Fable 5 is so powerful that Anthropic requires data storage for security monitoring
- Microsoft is concerned that insiders are sharing confidential information with Anthropic’s model
- Workers were told to use Microsoft alternatives, but third-party partner models are still available
Microsoft has reportedly restricted employees from using Claude’s recently launched Fable 5 model while legal and compliance teams review data handling.
According to Anthropic’s latest retention policy for the Mythos models, which includes the Fable 5, the company requires fast and output retention for security monitoring, which conflicts with enterprise-grade policies that customers like Microsoft have in place.
Anthropic stressed that prompts and output are required to be saved for 30 days, but content flagged by its security systems can be kept for up to two years if needed for investigation or enforcement.
Microsoft discontinues use of Claude Fable 5 due to data retention concerns
The current restriction, which applies to Microsoft employees, is in place because of concerns that insiders could share sensitive information such as customer information, company information and other confidential business information.
Fable 5 was launched after the huge success of Mythos, which was praised for its strong cyber security and vulnerability detection.
However, this model was only launched to a select group of partners due to concerns that it could be misused by attackers, hence the subsequent launch of the tamed-down Fable variant, which the company describes as “safe for general use.”
While likely unrelated, Microsoft’s push to halt Fable 5 deployment follows recent news that the company canceled internal Claude Code licenses and moved its developers to its own GitHub Copilot CLI tools.
Together, the two shifts reflect Microsoft’s desire to prioritize in-house tools over third-party alternatives, likely due to cost advantages, but a clear willingness to use systems from Anthropic also suggests the company is willing to lean on partners when necessary.
“It is not yet clear whether Microsoft’s legal teams will clear Claude Fable for internal use,” The edge added.
Even if a temporary ban is in place, internal demand may force Microsoft to change its tune. Mythos 5 and Fable 5 outperform key boundary models such as GPT-5.5, Gemini 3.1 Pro and Opus 4.8 across 13 test categories, including agent coding, knowledge work and cyber security.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
