Microsoft’s legal department is currently undertaking a thorough evaluation of Anthropic’s recently implemented data retention policy adjustments. These changes, specifically concerning the new Claude Fable 5 AI model, are prompting significant internal discussions.
Anthropic unveiled Claude Fable 5, its inaugural Mythos-class AI model, just yesterday, and it has already sparked apprehension within Microsoft. Reports indicate that Microsoft is imposing limitations on its employees' use of Claude Fable 5, directly attributable to Anthropic's updated data retention stipulations.
While Microsoft swiftly integrated Claude Fable 5 for its GitHub Copilot and Foundry clientele, the model is reportedly unavailable within the internal model selection interface used by Microsoft employees for their versions of GitHub Copilot. In contrast, all other Claude models remain accessible internally, operating under Zero Data Retention (ZDR) protocols.
Microsoft has informed its workforce that its legal teams are actively assessing Anthropic's revised data retention requirements. The primary areas of concern are understood to revolve around the handling of customer data and sensitive confidential information. It remains uncertain whether Microsoft's legal review will ultimately clear Claude Fable 5 for internal deployment.
Claude Fable 5 necessitates data retention for the proper functioning of Anthropic’s new safety classifiers. This policy dictates that Anthropic will retain user prompts and outputs, deleting this data after a 30-day period. However, certain prompts and outputs flagged for violating Anthropic’s usage policy could be stored for an extended duration of up to two years, raising complex legal questions regarding Microsoft’s appropriate use of Claude Fable 5.
As the first broad release from Anthropic’s advanced Mythos family of AI models, Claude Fable 5 emerges weeks after the company had indicated this model class was so proficient in cybersecurity tasks that it was deemed "too dangerous to release publicly." To mitigate these risks, Anthropic has integrated prompt safeguards, which are the underlying cause of these challenging data retention modifications.
Microsoft was contacted for comment regarding its utilization of Claude Fable 5, but the company did not provide a response prior to publication.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.