By using Fable 5 you understand that the model provider will retain agent request and output data associated with this model, regardless of your Cursor Privacy Mode setting.
This data will be used for automated and, where necessary, human harm prevention processes. Anthropic will never train on this data.
Comments
Quote from the dashboard: