Microsoft limits employee use of Anthropic's Claude Fable 5 over data retention concerns, The Verge reports
Microsoft is reviewing whether to approve the model for broader internal use as Anthropic’s new retention rules require prompts and outputs to be stored for 30 days.
- Microsoft is restricting employee access to Anthropic's Claude Fable due to data retention requirements, The Verge reported on Wednesday. The company's legal teams are currently evaluating the model's compliance with internal data policies.
- Anthropic mandates that prompts and outputs for its new Mythos-class models be retained for 30 days for trust and safety purposes. If flagged for usage policy violations, the startup may store this data for up to two years.
- Lawyer Jessica Eaves Mathews explained that this policy overrides existing Zero Data Retention agreements. Unlike other Claude models, Claude Fable lacks ZDR options, forcing a 'mandatory exception' for enterprise and API customers.
- Internal versions of GitHub Copilot currently exclude the model as Microsoft legal teams assess whether to approve it for broader use. Employees report the tool remains unavailable in the model picker.
- Researchers including Ellie Bakkouch criticized 'invisible safeguards' in Mythos models that intentionally limit performance on 'frontier llm research' tasks. These hidden interventions have sparked broader controversy regarding model transparency and developer control.
22 Articles
22 Articles
Microsoft's Quiet Caution on Claude Fable 5 Exposes AI Data Tensions
Microsoft sells Anthropic’s latest model to its biggest customers. Yet the software giant won’t let its own engineers touch it freely. Claude Fable 5 arrived this week with fanfare. Anthropic called it the first safe version of its powerful Mythos-class system. Benchmarks showed strong gains on complex reasoning and long-horizon tasks. Enterprises rushed to test it inside GitHub Copilot, Microsoft 365 Copilot and Azure’s Foundry platform. But a …
Microsoft has made the decision to restrict its workers to the use of Claude Fable 5, the first...
Zero Hedge
Nemos News NetworkMSFT Restricts Internal Use Of Claude Fable Over Data-Retention Concerns; BMO Calls Anthropic A Leading Pure-Play AI Lab
MSFT Restricts Internal Use Of Claude Fable Over Data-Retention Concerns; BMO Calls Anthropic A Leading Pure-Play AI Lab Anthropic released Claude Fable 5, a next-generation "Mythos-class" AI model, on Tuesday. The model is designed to restrict dangerous capabilities in areas such as cybersecurity and biological research after CEO Dario Amodei warned about risks last month. The model gives users access to Anthropic's more powerful Mythos model, …
Coverage Details
Bias Distribution
- 46% of the sources are Center, 45% of the sources lean Right
Factuality
To view factuality data please Upgrade to Premium
