AI Principles
Last Updated: September 9, 2026
These principles explain how to evaluate and use Venmail AI features, including their processing boundaries and limitations.
1. How processing works
- Venmail performs email analysis on its servers. Message content and extracted attachment text can be sent to Groq, a third-party AI provider, for analysis.
- AI features include summaries and extracting information that helps with follow-ups and scheduling. Availability depends on the feature and account.
- Managed plans currently use shared storage; regional selections are future preferences, not current data residency choices. Customer-provided storage does not by itself restrict where application or third-party AI processing occurs.
2. Review AI output
- AI output can be inaccurate, incomplete, or biased. Check summaries, dates, recipients, and action items against the original message before relying on them.
- Use human review for decisions that affect people, contractual obligations, or sensitive information.
- Report incorrect or unexpected results to hello@venmail.io with only the information needed to investigate.
3. Privacy and account choices
- Server-side AI analysis requires readable content. Venmail is not a blanket end-to-end encrypted or zero-access email service.
- Review our Privacy Policy and discuss the current subprocessors and account controls with our team before enabling a sensitive workflow.
- Do not assume a universal AI opt-out or a storage-region selection disables all content processing. Confirm which controls apply to the exact features you use.
4. Questions and feedback
Contact hello@venmail.io for current feature behavior, data processing questions, and AI feedback.