Content Policy

How this is enforced

Every generation request is screened automatically before it reaches a provider: a fast text check, a moderation model covering text and images, and a policy model. Requests that breach this policy are refused and recorded. Media submitted to the community gallery is reviewed before it appears publicly.

Never permitted

• Sexual content involving minors, or any sexualisation of a minor, real or generated. This is reported to the authorities where the law requires it.
• Non-consensual intimate imagery, including sexualised depictions of real identifiable people.
• Content that incites violence, terrorism, or hatred against a protected group.
• Impersonation of a real person to deceive, including synthetic voice or likeness used for fraud.
• Content designed to interfere with elections or mislead about voting.
• Instructions for weapons, explosives, or other material that would cause serious harm.

Restricted

• Depictions of real, identifiable people require their consent, and must not be presented as genuine.
• Graphic violence and adult content may be refused depending on the model and provider.
• Third-party trademarks and copyrighted characters remain the responsibility of whoever publishes the result.

Reporting and appeals

If you find content in the community gallery that breaches this policy, report it to [TO CONFIRM: privacy contact address]. If your request was refused and you believe that was wrong, contact us and a person will review it.

[TO CONFIRM: notice-and-action procedure and reporting obligations under the EU Digital Services Act, and CSAM reporting obligations in your jurisdiction.]