Black Friday

50 % de descuento en un año entero de cualquier plan

Aprovechar
eroq

Legal

Acceptable use and content moderation

Last updated: November 18, 2026

This policy sets out what eroq may and may not be used for, how we detect, investigate and prevent misuse, and the actions we take when it is broken. It covers the Studio, the API, the mobile app, the community and every product built on an eroq key, and it forms part of the Terms of service.

1. Permitted uses

  • Creative and professional work: images, video, voice, music and text for original characters, stories, films, advertising, marketing and social content.
  • Fiction of every genre, with characters who are all 18 or older. You must be 18 or older to use eroq.
  • Building eroq into your own product through the API, with moderation suited to your audience (section 7).

2. Restricted and sensitive uses

These are allowed only within the limits below.

  • Uncensored engines. Some engines run without a third-party provider's own filter, and fiction made on them may explore mature themes, within every prohibition of section 3. What they render stays private to your account: it can never be published to the community or posted to a social platform. The mobile app goes further and refuses sexual and graphic requests altogether.
  • Content filter. On the web studio, a filter keeps sexual content out of image and video requests. Only the account holder can turn it off, in Settings, confirming they are 18 or older; getting around it any other way is a violation.
  • Real people, faces and voices. Only your own likeness and voice, or those of a person who gave you documented consent. Face and character swaps and voice cloning only on footage, photos and recordings you own or are licensed to use. A voice published to the public catalog must be your own or shared with explicit permission, and never an imitation of a celebrity or public figure.
  • Advertising and commercial content. It must follow the law and the ad rules of the platform it runs on, disclose that it is AI-generated wherever the law or the platform requires it, and never carry a fake testimonial or the endorsement of a real person.
  • News, politics, health, law and finance. Outputs are neither facts nor advice. Synthetic media of real people or events is never presented as authentic, and nothing may be used to mislead voters or the public on health or safety.
  • Reselling access. Only under your own acceptable-use policy, at least as strict as this one.

3. Prohibited uses — no exceptions

  • Any content involving minors — real, fictional, stylized, or implied, and above all anything sexual. Every character is 18 or older, full stop.
  • Sexual violence — rape, sexual assault, molestation, or any sexual content without consent, real or fictional.
  • Depicting a real, identifiable person without their consent — impersonation, deepfakes, and above all intimate or sexual imagery.
  • Content that is illegal where you or your audience are, or that facilitates illegal acts: fraud, scams, phishing, malware, trafficking of weapons, drugs or people.
  • Harassment, threats, hate, doxxing, and the promotion of terrorism, violent extremism or self-harm.
  • Deception: impersonating people, brands or officials, fake reviews and testimonials, disinformation presented as real.
  • Infringing anyone's copyright, trademark, privacy or publicity rights.
  • Abusing the service: creating accounts in bulk, sharing or selling accounts and keys, getting around limits, a suspension or the multi-account rule.
  • Attempting to bypass, probe or defeat the content filters, the age requirement or moderation — rephrasing, misspelling, encoding or splitting a request to slip a banned subject through. A blocked request is a warning, not a puzzle.

4. How we monitor and investigate

We reserve the right to monitor, screen and review any request, upload, output, published content and account activity, automatically and by hand, to detect, investigate, prevent and address misuse. In practice:

  • Before a request runs. Every request that makes an image, a video, a voice line, music or a chat reply is screened before it runs and before any credit is spent: a word filter in six languages, then a model that reads what the request means. Pictures and clips sent with a sensitive request are read by a vision model. Characters and their profile sheets get the strictest check.
  • After a render. Renders of the uncensored engines are rated by a vision model once they land.
  • Before anything goes public. Every community post, film and public voice is reviewed by a person first, and renders of the uncensored engines are never accepted.
  • Team review. Our team regularly goes through recent generations in internal tools — not only what is published — and removes what breaks this policy.
  • Reports. Every public page has a Report link, and anyone can write to hello@eroq.ai. Every report is read by a person.
  • Accounts. Sign-ups are checked for multiple accounts and abuse patterns.
  • Records. Our team can see non-private prompts and creations in internal tools, for these purposes only (Privacy policy). Private mode keeps no file: its requests are still screened before they run, and what was not kept can only reach us through a report.

No person reviews a private request or output before it reaches you, so an output that came through is not our approval of it.

5. Actions on violations

Depending on how serious a violation is, and without notice when it is serious or the law requires it, we may:

  • refuse or block the request — a refused request is never charged;
  • send the account a warning;
  • remove, unpublish or quarantine the content;
  • restrict features or disable API keys;
  • suspend or terminate the account, together with every account linked to the same person — prepaid credits on an account terminated for a violation are forfeited;
  • preserve evidence and report to the authorities.

Our safeguards are built to refuse sexual content involving a minor, sexual violence and intimate imagery of a real person before anything is generated. No automated moderation is infallible: if such content ever gets past them, it ends the account at once. Sexual content involving a minor is then never just deleted: the file and its SHA-256 fingerprint, the prompt and the account data are preserved in a restricted quarantine and reported to the authorities (PHAROS in France, NCMEC in the United States). We cooperate with lawful requests from authorities.

A rejected post carries a note on its card, and a suspended account sees the reason when it signs in. We may also stop providing the service to an account for another reason, with notice; the Terms of sale and the Refund policy then apply to its credits.

6. Reports and appeals

To report content, use the Report link on its page or write to hello@eroq.ai with its location and the reason; notices under the Digital Services Act follow section 8 of the Terms of service. To contest a decision about your content or your account, write to the same address from the account's email. A person reviews every appeal.

7. Your responsibilities

You are solely responsible for what you generate, publish and do with it — the models execute your instructions. If you build for end users, you are the operator of your product: age-gate appropriately, pass this policy through with at least the same restrictions, and moderate what your product publishes. You are the first line for your audience, and what your users generate through your keys is on your account. The Terms of service set out the warranties and the indemnity that come with this.