Content Safety and Moderation

Overview

Video Lab screens prompt-based AI video generation requests before a job is created. The goal is to prevent prohibited content from reaching a generation provider and to keep paid credits tied only to allowed requests.

Moderation Workflow

For prompt-based video generation, the server follows this order:

1. Validate the request shape and required scene fields.
2. Screen the user prompt through the Creem Moderation API.
3. Continue only when the moderation decision is allow.
4. Block flag, deny, timeout, authentication error, malformed response, and other moderation failures.
5. Create a generation task, check credits, and call the video provider only after the prompt is allowed.

Blocked or flagged requests are not sent to the generation provider and do not consume generation credits.

Prohibited Content

Prompts, uploads, references, and outputs must comply with the Acceptable Use Policy. The service prohibits NSFW, adult, sexually explicit or sexually suggestive content, sexualized minors, non-consensual intimate content, deepfake or identity-misuse requests, harmful or illegal instructions, harassment, hate, graphic violence, and rights-infringing content.

User Responsibility

You are responsible for submitting only content you have the right to use. Content safety screening is an additional control; it does not replace your responsibility to follow the Terms of Service, Acceptable Use Policy, provider rules, platform rules, and applicable law.

Support

For safety, abuse, billing, or account questions, contact support@bernini.app.