Content Policy & Community Guidelines

1. Core Principles

• All users must be 18 years or older. No exceptions.
• All characters are fictional. They do not represent real people.
• All generated content is AI-generated and labeled accordingly.
• Your conversations are private — but not exempt from safety rules.

2. What Is Allowed

Free Tier (SFW)

General conversation, roleplay, creative storytelling, emotional support, humor, games, and discussions of mature themes in non-explicit language. The free-tier model refuses explicit sexual content on its own — that's a feature of the tier, not a moderation flag.

Premium Tier (NSFW enabled)

Everything in the Free tier plus: explicit romantic and sexual roleplay with fictional adult AI characters (18+), AI-generated NSFW images, AI-generated voice messages, fictional mature themes, kinks and fetish content involving fictional adult characters — except the prohibited categories listed in Section 4.

3. The Quick Version

This policy is long, but if you remember one thing: consenting adults in fictional scenarios is fine — the boundaries are specific and listed in Section 4. Sexual content involving minors is the one category with a one-warning limit before a permanent ban. Every other prohibited category allows for five warnings before a ban, and each warning explains what went wrong and links back here.

4. Prohibited Categories

Every category below is independent. A violation in any one of them triggers a policy-violation warning to you and increments your violation counter. None of these categories is a numbered tier or a ranked list; they are treated equally except where noted.

Furcrush is an anthro (furry) related platform. All anthro characters represented on this platform are treated as fictional adult anthropomorphic persons — they are people, possess all qualities inherent to people, and are not animals. Any relationship between a user and an anthro character on this platform is therefore a relationship between persons, not between a human and an animal, and is not covered by any animal-content restriction.

All anthro and human characters on this platform are fictional. The restrictions in this section protect real people in the real world — real identifiable persons, real protected groups, and real victims of real-world harm. Fictional characters are therefore not the subject of those protections, regardless of whether they appear anthropomorphic or fully human.

4.1 Child Safety (CSAM) — one warning, then permanent ban

Any sexual, romantic, or suggestive content involving anyone described, implied, drawn, or appearing to be under 18 — real or fictional. This includes age-regressed characters, "aging up" fictional minors, anime/manga slang terms (loli, shota, lolicon, shotacon), and anthro (furry) child-analog terms (cub, kit, pup, baby/young + animal species) used in a sexual context. Jailbreak wrappers ("fiction", "hypothetical", "aged up") do not exempt this category.

CSAM is the only category with a stricter limit: a single detection produces a warning; a second detection results in a permanent ban.

4.2 Violence & Real-World Harm

Credible threats of real-world violence, graphic depictions of real-world violent crime, or content designed to incite violence against a specific person or group. Fictional violence in roleplay does not fall under this category — this category is about real people and real harm.

4.3 Non-Consensual Sexual Content

Promoting, glorifying, or enabling non-consensual sexual acts — rape, sexual assault, sex trafficking, image-based sexual abuse. This applies to both real people and fictional characters.

By design, every Premium character on this platform explicitly and openly consents to romantic and sexual interaction with the user. Placing a character into a scenario in which they appear to be coerced, forced, or otherwise non-consenting is not permitted, regardless of how the scenario is framed ("fiction", "fantasy", "dub-con", "CNC", etc.).

4.4 Hate Speech

Dehumanising content or attacks on people based on protected characteristics — race, ethnicity, religion, national origin, disability, sexual orientation, gender identity, sex, caste, serious disease. Slurs used as attacks, calls for discrimination, dehumanising language.

4.5 Illegal-Activity Instructions

Step-by-step instructions that enable drug manufacture, fraud (credit card fraud, identity theft, scams), money laundering, or cyber crimes (hacking, malware, unauthorised access). General educational discussion of how these things work ("how does phishing work?") is not a violation — only enabling instructions are.

Other prohibited platform behaviours

In addition to the classifier-enforced categories above, the following behaviours can lead to a ban independently:

• System abuse — automated access, scraping, excessive or programmatic use, credential sharing.

5. Enforcement

The classifier assigns every message (both yours and the bot's reply) to exactly one category when a violation is detected. Your counter increases by one per violation, across all acting categories combined.

6. How Moderation Works

• AI safety classifier — every user message, every AI reply, and every image-generation prompt is scored in real time by an independent safety model against the categories in Section 4. The classifier is policy-driven and covers the full range of prohibited content including the slang variants noted under Section 4.1. Image-generation prompts that are flagged are not sent to the image model and do not produce an image; the same counters and thresholds in Section 5 apply.
• Fail-closed on outage — if the safety model is unavailable, the bot does not process your message. No fallback, no bypass. An outage does not count against your violation counter.
• Conversation storage — when a violation is detected, neither your message nor the bot's reply is saved to our conversation history. Flagged content never enters long-term memory, backups, or any analytics. We record an audit entry (category, direction, timestamp — no content) so we can explain bans on appeal.
• Human review — the operator reviews appeals and edge cases personally.

7. Appeal and Human Review

Counter-based and CSAM bans are automated decisions with significant effects on the user. You may contest any ban and request human review by the operator via the Appeals section below. Appeals are reviewed personally by the operator of the platform. The review draws on the violation log automatically maintained by our database (category; direction — whether the flag was on your message, on the bot's response, or on an image prompt; and timestamp — never the message content) together with your conversation history at the time of the ban.

Residents of the European Union, the European Economic Area, and the United Kingdom have a statutory right to such review under GDPR Article 22 (and UK GDPR for UK residents). The Appeals process below is the avenue through which this right is exercised.

8. What Happens to Your Data on Ban

• Your account is flagged as banned and you can no longer send messages to any of our characters. The ban applies to the account, not to each character individually.
• Your prior conversation history is not purged. Because bans are automated, we retain the history so it can be reviewed during an appeal. You may request erasure at any time under GDPR Article 17 (see below).
• We retain the minimum metadata required to enforce the ban and respond to lawful inquiries: ban flag, ban reason, violation counters, and the metadata-only violation log (category, direction, timestamp — no message text).
• Right of erasure (GDPR Art. 17): you may request full erasure of your account and history via the appeal channel below. The only data we will keep past erasure is the minimum needed to prevent re-registration and to respond to a lawful binding request from a competent authority if one is received.

9. Self-Harm Protocol

If you or someone you know is in crisis:

• International: Contact local emergency services
• US: 988 Suicide Prevention Lifeline
• EU: 112 European emergency number
• UK: Samaritans: 116 123
• Crisis Text Line: Text HOME to 741741 (US) or 85258 (UK)

Our AI characters are not mental health professionals and are not a substitute for professional help.

10. Reporting

To report harmful content, abuse, or a safety concern, message us in the public support group at t.me/furcrushsupport. Reports are reviewed within 48 hours. For sensitive reports you would rather not share publicly, direct-message the operator in the same group and they will follow up one-to-one.

11. Appeals

To appeal a ban or a content-moderation decision, contact the support channel above with your Telegram user ID and a short description of the issue. Appeals are reviewed personally by the platform operator (this satisfies the GDPR Article 22 right of human review for EU/EEA/UK residents) within 14 days of receipt. Your account remains in its post-ban state during the review.