Trust and safety
Responsible AI video, with receipts.
Guardrails, labels and a paper trail on every project.
Safety pre-check
Every prompt is checked before it reaches the video model. Requests for real people's likeness, sexual content, graphic violence and unsafe content involving minors are blocked, and never charged.
No real-person likeness
We don't recreate celebrities, public figures or private people. Original characters only, or people who've signed a release.
AI-content labels
Every clip carries Google's invisible SynthID watermark. Exports include AI-generated metadata so platforms can read it.
Voice consent
Custom voices need a consent check: your own voice, or written permission from the speaker. The record goes in your rights pack.
Rights pack
Each project keeps a paper trail of prompts, versions, model IDs, dates, consents and receipts, downloadable as CSV or PDF.
Business-grade models
We use generally available Veo 3.1 versions on Google Cloud, not preview models, and we leave Google's safety filters on.
Posting AI video? Label it.
TikTok, YouTube and Meta all label or ask you to disclose realistic AI-generated content, and some ad platforms require it before an ad can run. Use each platform's AI label when you post. We'll show reminders at export.
Your rights in the output come from our customer terms, which we'll publish before public launch. In the US, footage made only by AI may not be protected by copyright on its own; your script, edit and brand assets can be.
Make your first video with us.
Early-access spots open in small groups. Join the list or call 1-800-481-8638.