AI and your data
Veridox does not use your data to train AI. What each AI feature sends, and what Veridox keeps.
For: Everyone · Category: Safety & whistleblowing
Veridox does not use your data to train AI. AI features run through our infrastructure provider's secure AI gateway (InsForge), and every request instructs the AI providers not to keep or train on your data. Veridox keeps only the results you see in your workspace, such as summaries, suggestions and alerts. When the AI safety scan flags a message, the flagged message text is kept with the alert so that owners and admins can review it.
Steps
- Where AI is used: Dashboard suggestions, your personal activity summary, chat translation, call summaries (only where call recording and transcription are on) and the AI safety scan (only where an owner or admin turns it on).
- How a request is sent: Our servers send each AI request through the Model Gateway of our infrastructure provider, InsForge, which routes it through OpenRouter to the model provider. Every request asks only for model providers that do not keep the data or train on it, on zero-data-retention endpoints. If none is available, the request is refused and the feature does not run.
- What Veridox keeps: Only the results you see: dashboard suggestions, your activity summary and call summaries. Chat translations are shown on your screen and are not saved by Veridox. When the AI safety scan flags a message, its text is kept with the alert: owners and admins can read it, and HR sees only that an alert exists.
- Your controls: Owners and admins can switch dashboard suggestions off for the whole workspace in Settings. The AI safety scan stays off until an owner or admin turns it on. During the free trial each workspace gets 100 AI uses a week.
Tips
- Full details are in the Privacy Policy (section 5) and the Data Processing Addendum.
- Questions about your data: privacy@veridoxspace.com.