Trust & Safety
VidRocket makes it easy to generate realistic images, video, and voices. That is the point of the product, and it is also why we take misuse seriously. This page explains what we do to keep the Service safe, what we expect from you, and what happens when something goes wrong. The rules themselves live in the Acceptable Use Policy; this page is about how they are enforced.
1. Our commitments
- No training on your content. We do not use what you upload or make to train our own models, and we do not enable training at the providers we use.
- Consent before processing. The apps ask for explicit consent before your media is sent to an AI provider, and every account accepts the content responsibility notice once before creating anything.
- Provider safety checks stay on. We do not disable or work around the safety filters our model providers run, and we do not let users do so either.
- Humans in the loop. Automated screening flags; a person decides on takedowns and account actions.
2. How we screen content
- At upload time, every photo you bring in — a product shot, a creator photo, a reference, a picture dropped into chat — is checked automatically before it is stored. A photo that breaks the Acceptable Use Policy is refused and deleted rather than kept, so it never reaches your library. Nothing is charged for a refused upload.
- At generation time, the model providers apply their own filters and may refuse a request that appears to involve a real person, a public figure, a minor, or prohibited subject matter. We pass those refusals through and refund the credits.
- At publish time, every item published to the public Gallery or the community feed is screened automatically before it is shown, and again when it is reported. Items that fail screening never go live.
- On UGC and talking-creator takes, the finished clip goes through a frame and transcript check so that what was generated matches what was asked for.
- Private content is not proactively reviewed by people. We look at a private item only when it is reported, when a provider flags it, or when the law requires us to.
3. Real people and consent
The most serious misuse of a tool like this is putting someone’s face or voice into media they never agreed to. We treat these reports as the highest priority:
- If you are the person depicted and you did not consent, email support@vidrocket.ai with the link or a description of the content. You do not need a copyright claim; your likeness is enough.
- We aim to act on likeness reports within one business day, and we remove the content first and sort out the details after.
- Accounts that upload other people’s faces or voices without permission lose generative features on the first confirmed report and lose the account on the second.
- Intimate or sexual imagery of a real person without consent, and any sexualized depiction of a minor, results in immediate account termination and, where the law requires it, a report to the relevant authority. We preserve the evidence.
4. Reporting
You can report from inside the product or by email:
- Public Gallery or community item: open the item and choose Report.
- A creator profile, shared analysis, or board: use the Report link on the page.
- Your own generation went wrong (offensive or unexpected output): open the item in Assets, choose the menu, then Report. That tells us and the provider.
- Anything else: support@vidrocket.ai. For copyright and trademark, use the process in the Copyright & Trademark Policy.
Include a link where you can. We review every report and reply when we have taken action or decided not to. We do not share the reporter’s identity with the account reported.
5. Enforcement
We match the response to the harm. In rough order of severity:
- Refusal. A generation is declined before it runs. Credits are not charged.
- Removal. An item is taken down from the Gallery, feed, or a shared link, and you are told why.
- Feature limits. Generative features that involve real people (creators, voices, animation from photos) are switched off for the account.
- Suspension. The account is paused while we investigate.
- Termination. The account is closed and cannot be reopened. Credits and subscription time are forfeited where the closure is for a policy breach.
Some breaches skip straight to termination: anything involving minors, non-consensual intimate imagery, terrorism, or attempts to defeat provider safety checks.
6. Appeals
If we removed something or limited your account and you think we got it wrong, reply to the notice we sent or email support@vidrocket.ai with the subject “Appeal”. Tell us what was removed and why you believe it was allowed. A different person from the one who made the original decision reviews it, and we answer within seven business days. Decisions about content involving minors are not appealable.
7. Law enforcement and legal requests
We respond to valid legal process from the jurisdictions where we operate and, where we are legally allowed to, we tell the affected user. We preserve content and account data when a preservation request or a serious-harm report requires it. Emergency requests involving an imminent risk to life go to legal@vidrocket.ai with “Emergency” in the subject.
8. Labeling AI-generated media
Everything the generative features make is AI-generated, and you should treat it that way when you post it. Our publishing tools set the AI-content disclosure on platforms that support one. Where a model provider adds a watermark or provenance metadata, we leave it in place. Laws on synthetic-media labeling vary and are changing quickly; you are responsible for the rules where you post.
9. Security and vulnerability reports
Found a security problem? Email legal@vidrocket.ai with the details and give us a reasonable window to fix it before publishing. We will not take action against good-faith research that stays within the Service you are entitled to use and does not touch other users’ data.
10. Contact
Safety reports and appeals: support@vidrocket.ai
Legal, emergency, and security: legal@vidrocket.ai