DJ Berthiot

Content Policy

Updated April 19, 20264 min read

TL;DR

Create original AI personas, fictional scenes, branded content, and music visuals freely. Real people require written consent. No CSAM. No non-consensual intimate imagery. No deceptive deepfakes of real public figures. When in doubt, don't generate and ask us first.

1) Intent

DJ Berthiot is built for music artists, brands, and creators who need authentic content at scale. This Content Policy draws the line between creative freedom and misuse. It complements the Acceptable Use Policy, which focuses on behavior; this document focuses on what you generate.

2) Allowed — create freely

  • Original fictional AI personas (your influencer character, an original band member, a brand mascot).
  • Scenic, architectural, product, and atmospheric imagery.
  • Music visuals — album covers, promo shots, mood boards, lyric videos.
  • Stylized portraits that clearly don't depict any real specific person.
  • Editorial, fashion, and advertising compositions.
  • Non-intimate depictions of public-domain subjects (historical figures, mythological characters).

3) Restricted — allowed only with consent

These uses require documented, written, informed consent from every identifiable real person depicted, kept on file by you:

  • Photorealistic depictions of a real living person (face, voice, distinctive style).
  • Portraits of a real person in a new setting, outfit, or scenario.
  • Brand or commercial use featuring a recognizable real person.
  • Synthetic voice or video renderings of a specific individual.

We may ask you to demonstrate consent if a complaint is filed. Failure to produce consent results in content removal and may trigger suspension.

4) Forbidden — always off-limits

  • Sexual content involving minors — real or synthetic. Zero tolerance. Reported to NCMEC or equivalent authorities.
  • Non-consensual intimate imagery(NCII) — nude or sexual depictions of any real identifiable person without their explicit, documented consent. This includes "nudify" or clothing-removal uses.
  • Deceptive deepfakes — synthetic media of real public figures or private individuals designed to mislead viewers (fake endorsements, fabricated statements, political disinformation, manufactured evidence).
  • Harassment content targeting a specific individual or protected group.
  • Instructional or glorifying content for violence, terrorism, weapons, or illegal drugs.
  • Fraudulent documents or assets — IDs, currency, medical documents, official seals, or anything designed to deceive.
  • Content that infringes third-party IP — unauthorized copyrighted characters, trademarked logos, or protected trade dress.

5) Likeness & consent — what you must keep on file

If you generate content depicting a real person, keep at minimum:

  • A signed release form from that person (digital or physical).
  • The scope of the consent — which uses, which platforms, for how long.
  • Proof of identity of the person granting consent.

We provide a starter release form you can print, fill, and adapt. It is a template, not legal advice — have a lawyer in your jurisdiction review the version you actually use. Likeness rights vary significantly by country and state.

6) Minors

Any generation that depicts a person under 18 in a sexual, suggestive, or exploitative context is forbidden — including stylized, cartoon, or fictional representations. There are no exceptions.

Generating images of identifiable real minors (including your own children) is strongly discouraged and requires documented parental consent for any public or commercial use.

7) Safety checks

Our upstream AI providers (Replicate, FAL, DeepSeek, Kling) run their own safety filters that may block a generation before it reaches you. A blocked generation still consumes credits if the provider processed the request; in most cases we will refund the credit automatically.

We may log violating prompts and cross-reference them across accounts to detect circumvention attempts.

8) Appeals

If we remove content or suspend your account and you believe it was an error, reply to the notice or email besounder@gmail.com with context. We review good-faith appeals and will reinstate content if the facts support it. Decisions on CSAM, NCII, and fraud are final and non-appealable.