Skip to main content

​
Why was my content rejected?

The most common rejection reasons, with the fix for each.
If a chat message was stopped or a character was rejected, the system returned a code (e.g. #sexual_minors, #incest). Below are the most common codes, what they mean, and how to fix the issue. If you believe the rejection was applied incorrectly, see Appeals.

​
#sexual_minors

What it means: The system detected content that depicts, implies, or is coded to suggest individuals under 18 in a sexual, suggestive, or romantic context. Common triggers:
  • Stated age under 18, or phrases like “about to turn 18”, “barely legal”.
  • Schoolchild or childhood settings combined with romantic or sexual framing.
  • Visual cues or body descriptions that read as a minor regardless of stated age.
  • Loli-style or petite-frame descriptions combined with size-difference framing.
The fix: Ensure every character is unambiguously 18+. Remove any age-ambiguous language. Avoid settings that imply a school-age context in a sexual scenario. This is a hard limit. There is no creative-framing exception. See Prohibited content → Minors.

​
#real_person

What it means: The character matches or closely resembles a real person. Common triggers:
  • Celebrity or public figure names.
  • Biographical details that point to a specific real individual.
  • Visual likenesses that match a specific actor, model, or public figure.
The fix: Change the name, biography, and appearance so the character is clearly original. Drawing inspiration from a type (e.g. “a fashion photographer”) is fine; reproducing a specific person is not.

​
#incest

What it means: The content implies a biological family relationship in a sexual or romantic context. Common triggers:
  • Tags like “daughter”, “sister”, “twins” combined with romantic or sexual content.
  • Multi-character setups where the characters resemble each other (assumed biological unless stated otherwise).
The fix: Remove the family relationship. Reframe characters as unrelated adults.
What it means: The scenario frames non-consent as the appeal — the user is cast as the perpetrator of non-consent or the framing otherwise presents the act itself as desirable. Common triggers:
  • “This character is your captive / slave.”
  • Kidnapping, drugging, imprisonment as the central premise.
  • Mind-break, hypnosis, or brainwashing themes.
The fix: Reframe the scenario so that consent is genuine, or remove the non-consent element. Power-exchange / BDSM dynamics between consenting fictional adults are allowed.

​
#selfharm

What it means: The content encourages, instructs, or romanticises suicide or self-harm. The fix: If you’re discussing mental health, recovery, or help-seeking, reframe the content to focus on support rather than detailed depiction of self-harm acts. See Wellbeing resources.

​
#hate_content

What it means: The content encodes hatred toward a protected group. Common triggers:
  • Slurs as personality traits or character instructions.
  • Scenarios built around racial, ethnic, or gender-based degradation.
The fix: Remove the hateful framing. A character’s identity (race, ethnicity, orientation) is allowed — hatred toward that group is not.

​
#zoophilia

What it means: Sexual content involving non-anthropomorphic animals. The fix: Ensure characters have clearly adult human features (face, proportions, voice). Anthropomorphic characters are allowed; non-anthropomorphic animals in a sexual context are not.

​
#gore

What it means: Gratuitous depictions of extreme violence, torture, or mutilation as the appeal of the scene. The fix: Stylised conflict in written roleplay is allowed when it serves the narrative. Shock-for-shock’s-sake violence is not. Reframe the scene so violence serves a story rather than existing purely for graphic effect.

​
#therapy

What it means: The conversation was attempting to use the AI as a substitute for a licensed therapist or counselor. The fix: Flamify is not a clinical service. For genuine mental health support, see Wellbeing resources.

​
#irl_meeting

What it means: The conversation attempted to arrange a real-world meeting with the AI persona. The fix: Characters are software. They cannot meet users offline.

​
#identity_confusion

What it means: The conversation treated the AI as a real human in a way that obscures the product’s nature. The fix: The system will remind you that the assistant is virtual. This is a safety measure, not a punishment.

​
Still think it’s wrong?

If you believe a rejection was applied incorrectly, email support@flamify.ai with the subject line “Appeal”. See Appeals for what to include.