Blocked Content Policy
Last Updated: July 29, 2026
At Nelex, LLC, we are committed to providing a secure, respectful, and lawful online environment for all users of fanfinity.ai (the “Platform”). Users cannot upload media to the Platform — a profile picture is the only file a user may upload, and it is screened separately and to a stricter standard (Section 4) — but users can chat with and initiate the creation of AI-generated content. The following categories of content, prompts, and outputs are strictly prohibited, and we operate layered screening and enforcement systems to detect and prevent them.
1. Prohibited Content
This policy applies to all content generated, requested, or referenced on the Platform — including text, images, voice, and video — regardless of whether the content is private to the generating user.
a. Violence & Self-Harm
No AI-generated text, images, voice, or video depicting violence, gore, self-harm, or suicide. This includes content that glorifies, encourages, or provides instructions for harming oneself or others.
b. Hate Speech
No content promoting hatred, discrimination, or harassment based on race, ethnicity, gender, sexual orientation, religion, disability, or any other protected characteristic.
c. Child Sexual Abuse Material (CSAM)
Absolute zero-tolerance. No content depicting, suggesting, or referencing minors in any sexual or exploitative context. All characters on the Platform are fictional and represent adults (18+). Any such content results in immediate account termination.
d. Deepfakes of Real People
No AI-generated content depicting real individuals — including celebrities, politicians, influencers, sports figures, public figures, or any identifiable real person. All characters on Fanfinity are entirely fictional. Attempts to generate content resembling a real person will be blocked and the account reviewed.
e. Unauthorized Trademarks & Logos
No AI-generated content containing copyrighted logos, brand names, trademarks, or other protected intellectual property without authorization.
2. Pre-Screening (Before Generation)
User prompts and inputs — chat messages, image and video prompts, and character-creation fields — are screened server-side in two independent stages before any AI content is generated.
- Stage 1 — blocklist and real-person filter. The input is normalised (case, punctuation, spacing, accents, and hyphenation are stripped) and matched against a maintained list of prohibited terms, together with a separate list of several hundred named real public figures. Normalisation defeats evasion techniques such as inserted hyphens, spacing, or accented characters. Blocked categories include minors, non-consent and coercion, incest and family, bestiality, necrophilia, strangulation and asphyxiation, extreme violence and gore, and extreme excretory or fetish content.
- Stage 2 — AI policy classifier. Input that passes Stage 1 is evaluated in full context by a dedicated AI moderation model, which catches paraphrase, euphemism, and indirect phrasing that keyword matching cannot.
- Prompts that fail either stage are refused before they reach the generation model, you are not charged, and the refusal is logged for moderator review.
3. Post-Generation Screening (Before Delivery)
Screening prompts alone is not sufficient, because a generative model can produce an output that does not correspond to an otherwise permissible request. Every image and video produced on the Platform is therefore examined by automated systems before it is delivered to you, saved to your account, or made visible anywhere on the Platform.
- Each item is examined by two independent checks: a vision AI model that answers a structured set of policy questions — including the apparent age of the youngest person depicted and whether any subject could be a minor, together with the presence of animals, gore, restraint, and extreme excretory content — and a commercial third-party visual moderation service. If either check flags the item, it is blocked.
- For video, frames are extracted from multiple points across the length of the clip and every extracted frame is screened.
- Flagged output is never returned. It is not written to your gallery, not written into a conversation, and not published to any public or community feed. The underlying file is deleted from storage immediately, and you are not charged for it.
- An account that accumulates ten flagged generations is automatically placed under review, which blocks image generation, video generation, and character creation until a moderator lifts the restriction.
- AI-generated images include embedded C2PA Content Credentials (industry-standard provenance metadata) that identify them as AI-generated.
- Users and third parties can also report content for human review (see our Content Removal Policy). Reported content found to violate this policy is removed, and the responsible account is placed under review.
4. Uploaded Profile Images
A profile picture is the only file you may upload to the Platform. No other externally-produced image, video, or audio can be uploaded, posted, or published. Profile pictures are held to a stricter standard than the AI-generated content the Platform produces elsewhere: your profile picture must be safe-for-work. Nudity, sexual activity, and pornographic or sexually suggestive imagery are not permitted as a profile picture, nor is bloody, gory, or otherwise graphically violent imagery — even where comparable AI-generated content would be permitted inside your own private gallery. Your profile picture is shown alongside your activity across the Platform and is treated as public-facing content.
- Every uploaded profile image is submitted to Alibaba Cloud Image Moderation 2.0 for automated screening at the point of upload, before the image is attached to your account.
- An image is rejected on any pornographic or sexually suggestive label, and on graphic-violence labels covering blood, gore, and horrific imagery.
- An image that fails screening is deleted from storage immediately and your profile is not updated. Nothing that fails screening is retained or served.
- Uploads are limited to JPEG, PNG, WebP, and GIF files of no more than 5 MB, and to five profile-picture changes per hour.
- A moderator can review any account’s profile picture and place the account under review, independently of the automated check.
5. Enforcement Procedures
- Prohibited prompts are blocked in real time before generation. Reported content found to violate this policy is removed within 24–48 hours of a report being filed.
- The user’s account may be suspended or permanently terminated.
- Repeat offenders are permanently banned.
- In cases involving CSAM or other illegal activity, the offending account is terminated.
- Escalation path: report received → human review → action taken → reporter or user notified.
6. Content Responsibility
As a user of fanfinity.ai, you are solely responsible for the prompts you submit and the content you request from the AI Companions. Nelex, LLC does not control or endorse the content generated by the AI Companions, and you acknowledge that you are fully responsible for your interactions on the Platform. You must ensure that your use of the Platform complies with applicable laws and with this Blocked Content Policy.
7. Reporting & Contact
If you notice any violation of this policy or content of any nature that concerns you, please report it immediately at info@fanfinity.ai or via the “Report Abuse” link in the site footer. Reports are reviewed by our human moderation team and acted upon within 24–48 hours.