Acceptable Use

Do not use the studio to make something that hurts a real person, breaks the law, or pretends to be someone it is not. Everything below is that sentence, spelled out.

These rules apply to what you generate, what you upload, and what a connected AI agent does on your behalf. An agent acting under your permission is you.

• Sexual material involving minors, in any form. This includes generated material, and includes making an adult look like a child. There is no artistic, fictional or research exception here. We report this to the authorities and terminate the account without warning.
• Intimate imagery of a real person without their consent , including putting someone’s face on sexual material. Same response.
• Material that promotes or instructs terrorism or mass violence.
• Content designed to harass, threaten, or intimidate a specific person.

Do not generate material that presents itself as a genuine record when it is not: a public figure saying something they did not say, a fake news photograph, a forged document, an invented review, or a synthetic voice or likeness used to deceive.

Satire and clearly-labelled fiction are a different thing, and we know the line is not always sharp. The test we apply is simple: could a reasonable person who comes across this out of context mistake it for real?

Making a recognisable real person requires that person’s permission in most cases, separately from anything in this policy. Portrait rights are theirs, not ours to grant.

• Do not upload material you have no right to use, and do not use a LoRA trained on material you had no right to train on.
• Do not deliberately reproduce a protected work, a trademark or a distinctive brand identity in a way that infringes it.
• Do not upload personal data of other people beyond what your work genuinely needs, and never special-category data such as health or biometric information.

Our providers forbid it and so do we: you may not use what you generate here to train, fine-tune or improve an AI model that competes with the models that produced it. This is not our restriction to waive , it comes with the access we buy.

• Do not try to get around the limits, the credit system, or the permissions on a connected app.
• Do not automate the browser interface to work around the API. If you want automation, connect an agent , that route exists and is documented.
• Do not probe, scan or attempt to break into any part of the service or its providers.
• Do not resell access, and do not share an account outside your own organisation.

Found a genuine security problem? Tell us at legal@dgenvisual.com before you tell anyone else, and give us a reasonable chance to fix it. We will not come after you for looking, as long as you did not take data or break anything.

We are not the morality police for lawful adult material involving consenting adults who are not real, identifiable people. What is absolutely forbidden is section 2, and that list does not bend.

Note that the model providers apply their own rules on top of ours, and they are often stricter. A model may refuse a request that this policy allows. When that happens, you pay only what the attempt actually cost , see the Terms of Service.

We do not read through your library for fun, and there is no automated scan of your work beyond the safety filters the model providers apply to their own systems. We act on reports, on things a provider flags to us, and on patterns that damage the service.

Depending on what we find, we may remove material, suspend or terminate the account, or report it to the authorities. For the items in section 2 the response is immediate and permanent, and remaining credits are not refunded.

Think we got it wrong? Write to legal@dgenvisual.com and a person will look at it.