Rules for AI Companion Communities: A Practical Moderation Template

Define the subject and audience first

AI companion communities often contain several different kinds of discussion. Members may talk about fictional characters, conversational features, storytelling, platform design, personal reactions, or the wider social questions raised by artificial companions. Without a clear scope, moderators end up deciding whether every post belongs through individual arguments.

Begin the rules by stating what the community exists to discuss. For example: This community is for discussion of AI companion platforms, fictional AI characters, conversational experiences, creative uses, and the social issues surrounding these services. Posts should contribute to one of these subjects and follow the channel-specific rules.

Set an explicit audience boundary. Some services and topics are intended for adults, while a community may choose to operate with a different age policy of its own. State the minimum age for participation and apply the relevant platform requirements. An adult-oriented subject does not automatically mean that every discussion belongs in every channel.

For example, TheFriend is an adults-only (18+) AI companion platform where every character is a fictional AI character. A community that discusses it should apply its own age-appropriate channel rules and avoid treating discussion of a fictional character as permission to ignore the community's content restrictions.

A copy-ready foundation for community rules

The following template gives volunteer moderators a starting point. Replace bracketed details with the community's actual choices, then remove provisions that do not fit the platform or audience. A rule that describes an action the team cannot enforce will create arguments without improving moderation.

• 1. Scope: Posts must relate to AI companion platforms, fictional AI characters, conversational experiences, creative projects, or relevant community discussion.
• 2. Age and access: Members must meet the community's minimum age requirement of [age]. Content must follow the age and content restrictions for its assigned channel.
• 3. Respectful participation: Criticize platforms, characters, features, and ideas without targeting members with insults, harassment, threats, or repeated unwanted contact.
• 4. Personal experiences: Discussion of emotional reactions and personal experiences is permitted when relevant and respectful. This community is not a crisis service or a substitute for professional support.
• 5. Promotion: Advertising, referral links, unsolicited invitations, disguised endorsements, and repetitive promotional posts require prior moderator approval or are prohibited.
• 6. Privacy and consent: Do not publish another person's identifying information, private communications, or material shared without appropriate permission.
• 7. Content organization: Use the required tags and channels, provide relevant context, and follow restrictions on explicit, disturbing, or otherwise sensitive material.
• 8. Enforcement: Moderators may remove content, restrict participation, or suspend accounts for rule violations. Members may appeal moderation decisions through [appeal method].

Separate personal discussion from off-topic support requests

Emotional discussion is one of the areas most likely to produce inconsistent enforcement. A member may describe feeling attached to a fictional character, frustration with a conversation, or disappointment after a platform change. These posts can be relevant to a community about AI companions. Treating every personal reaction as off-topic would remove a substantial part of the subject.

The useful distinction is between discussion that contributes to the community and requests that require a different kind of support. A post asking how members understand attachment to fictional characters is relevant. A post asking the community to provide continuous crisis support, diagnose a condition, or take responsibility for someone's immediate safety requires a different response and may need to be redirected.

Write the boundary plainly. For example: Personal experiences are welcome when they relate to AI companion use or community discussion. Posts seeking urgent personal support, professional diagnosis, or ongoing one-to-one emotional care may be redirected or removed. Moderators are volunteers and cannot provide crisis intervention.

Give members a reasonable alternative where one exists, such as a designated discussion channel for general personal reflections. Avoid turning the rule into a judgment about whether someone's feelings are legitimate. Moderators can enforce scope and conduct without evaluating the emotional importance of an experience.

Make promotion rules specific enough to apply consistently

Communities about AI companion platforms naturally attract references to products and services. Mentioning a platform is not automatically advertising. A member may compare character design, discuss an interface, ask about a feature, or describe a problem. A blanket ban on naming services would make ordinary discussion difficult to moderate.

Define promotional behavior by its purpose and presentation. Examples include referral links, repeated calls to join a service, unsolicited discount announcements, coordinated promotional posts, and accounts that primarily exist to advertise. If platform owners or representatives are allowed to participate, specify whether they must disclose their affiliation and where they may post.

A practical rule could state: Discussion of platforms and services is permitted when relevant to the topic. Promotional links, affiliate material, recruitment, and repeated advertisements require prior approval. Moderators may remove posts that primarily seek traffic, sign-ups, or commercial engagement.

The failure mode is treating enthusiasm as proof of promotion. A member who likes a service may write positively about it without being affiliated with its operator. Apply the same criteria across favorable and critical posts, and ask for clarification when an account's relationship to a service materially affects the discussion.

Organize channels around moderation needs

Channel design can prevent disputes before a moderator sees a report. A subreddit might use post flairs for platform discussion, character discussion, creative work, personal experiences, and policy questions. A traditional forum can organize equivalent subjects into separate boards. A Discord moderation setup can use dedicated channels with descriptions and posting permissions.

Start with a small structure. For example, a community might have general discussion, platform-specific discussion, creative content, and personal experiences. Add an adult-only or restricted-content area only if the platform's rules, community age policy, and moderation capacity support it. Too many channels create empty spaces and make it harder for members to understand where a post belongs.

Write a one-sentence purpose for every channel. State whether links are allowed, whether content warnings are required, and whether members may share screenshots or creative material under the community's general rules. Verify that the available platform controls can actually enforce the access restrictions being advertised.

Avoid using channel separation as a substitute for a clear conduct policy. Harassment, threats, and privacy violations remain prohibited across the server or forum, including in channels that allow more personal discussion.

Use volunteer moderator tools for narrow, repeatable tasks

Automated moderation can help manage recurring administrative work. A spam filter can identify repeated promotional links, an account-age rule can hold selected posts for review, and a bot can remind members to select a required topic tag. These functions reduce repetitive queue work when their triggers are narrow and visible to moderators.

A subreddit moderation workflow might use automatic removal for a known spam pattern, a temporary hold for links from new accounts, and a moderator review queue for posts that contain sensitive terms. In a forum, similar controls may be available through built-in moderation settings. Chat servers can use role permissions, channel-specific posting restrictions, and moderation bots, depending on the platform and its current features.

Do not let a keyword filter decide whether a personal discussion is appropriate simply because it mentions loneliness, attachment, or distress. Such terms can appear in thoughtful policy discussions and ordinary conversation. The tool should identify content for human attention when context matters.

Give every automated action a failure procedure. A moderator should be able to inspect the original content, restore an incorrectly removed post, and adjust the triggering rule. Begin with alerts or review holds where practical. Expand automatic enforcement only after the team has a clear understanding of the consequences and can manage appeals.

Define enforcement and appeals before the first dispute

Rules need an enforcement process that explains what happens after a violation. A basic sequence might include a reminder, content removal, a temporary restriction, and a longer suspension for repeated or serious behavior. The sequence should allow moderators to respond proportionately to the severity of an incident rather than mechanically issuing every step.

State that moderators may skip intermediate steps for serious conduct, including threats, targeted harassment, privacy violations, or deliberate attempts to disrupt the community. Also explain how members can appeal, which moderator reviews an appeal, and whether the original decision may be changed.

For volunteer teams, record the reason for consequential actions in a consistent format. A short internal note should identify the rule, relevant content, action, and reviewer. Members should receive enough information to understand the decision without being given private details about other accounts or internal security measures.

The final step is to publish the same core rules wherever members encounter the community. Assign one moderator to compare the subreddit rules, forum policy, server welcome message, and automated enforcement settings. Update the documents together whenever the age boundary, channel structure, or enforcement process changes.

Questions readers ask