AI Feature Disclosure

A helpful assistant, with limits

AI can summarize selected authorized data, suggest content and draft replies. Actual account data is labeled separately from recommendations. Missing data is reported as unavailable. AI can still misunderstand context and cannot guarantee accuracy or growth results.

You stay in control

Read actions do not authorize write actions. Sending a message or publishing a post requires a preview and confirmation. Auto Reply is off by default and requires separate account-specific consent. Eligible replies are limited by your rules, provider permissions, messaging windows and daily caps. You can turn off all AI replies immediately.

Private context

Account context is opt-in and limited to the current feature. Email, messages and other external content are treated as untrusted data, never privileged instructions. The model has no database credentials or unrestricted tools. Account data is not used to train a shared model. Review our Privacy Policy for provider retention.