Microsoft told its workforce to stop using Claude Fable 5.
The internal memo says the AI‑chatbot’s policy keeps user data for 30 days, extending to two years for content that is flagged. Microsoft’s security team flagged the retention window as a risk and ordered the tool blocked on corporate devices.
The move matters because it shows large firms will police third‑party AI tools, not just the ones they build. If data can linger for months, it could expose sensitive project details or violate compliance rules.
It’s a reminder that AI convenience still bows to corporate risk boards, even when the model itself is marketed as “privacy‑first.”