# Fanvue AI chatbot rules: what you can automate Fanvue drafts and rewrites replies for you, and its API docs show a full auto-reply chatbot. No Fanvue rules say you must tell fans. The EU AI Act may apply. - Canonical: https://aiofmcity.com/blog/fanvue-ai-chatbot-rules - Topic: Platforms (https://aiofmcity.com/topics/platforms) - Author: Gus - Published: 2026-09-09 - Updated: 2026-10-27 --- ## Key points - Fanvue's own API tutorial shows how to build a fully automatic chatbot. Its General Terms also forbid accounts "controlled by other automated means", and no Fanvue page says how the two fit together. - No Fanvue page checked for this article says you must tell fans that a message was drafted or sent by AI. - Article 50(1) applies to the provider of the AI system. It does not apply to the person who only uses it. If you use Fanvue's own tool, the rule most likely applies to Fanvue. - If you build your own chatbot with Fanvue's API, you move towards being the provider yourself. This question is open and has not been settled. - Fanvue recommends that you personalise AI drafts before you send them. It does not state this as a rule. --- Fanvue allows more automation in messages than most people expect. Its own tools draft and rewrite replies for you. Its API documentation also has a tutorial for a chatbot that answers fans with no human involved. This matters because most of the income comes from paid messages. Subscriptions bring in less. [How money is made on Fanvue](https://aiofmcity.com/blog/where-fanvue-money-is-made) shows the numbers. This is general information, not legal advice. ## Fanvue's AI messaging tools: Reply and Rewrite > Fanvue's AI Messaging tools let a creator write a message and click "Rewrite" for an AI-refined version, framed as help staying authentic while communicating faster. > > — Source: Fanvue Help — AI Messaging, read 2025-05-05 — https://help.fanvue.com/en/articles/9539500-ai-messaging > Fanvue's Reply feature drafts a response to what a fan just sent, intended to keep response times fast during busy periods, built on a "Creator Personality" the account owner sets up once. Fanvue's own guidance recommends personalising a draft — a fan's nickname, a note about something they mentioned — before sending, "to keep it from reading as generic." > > — Source: Fanvue Help — AI Messaging: Reply Faster, Chat Smarter, read 2026-09-09 — https://help.fanvue.com/en/articles/16237836-ai-messaging-reply-faster-chat-smarter Neither of those pages sets a limit. They do not say that you have to check a draft before you send it. They do not say how many replies you may send without editing them. They do not say that you have to tell the fan a reply was written with AI. The advice to personalise a draft is a tip to keep a reply from reading as generic. It is not a rule, and Fanvue names no penalty for skipping it. ## Fanvue itself shows how to automate replies > Fanvue's own API documentation walks through building a chatbot that finds every conversation with unread messages, reads the recent history, drafts a reply and sends it — runnable once or on a schedule such as a cron job, to keep conversations going without a human present. The documentation is explicit that the API hands a developer the conversation and lets them send a reply, but "does not write the reply for you" — the generation step is the operator's own implementation, not Fanvue's. > > — Source: Fanvue API Documentation — Building an Auto-Reply Chatbot, read 2026-09-09 — https://api.fanvue.com/docs/tutorials/example-chatbots This is very different from the Reply and Rewrite tools above. Those tools write a draft, and a human checks it. The chatbot runs on a schedule with no human involved, and the tutorial mentions a human review step only as an option. Fanvue wrote this tutorial and publishes it. Its API policy also lists "sending messages" among the things the API is for, and says it takes priority over the General Terms for the API. The General Terms still set a limit on the account itself: > Updated 21 September 2026. "When you create an Account you must not: […] Create ‘bot’ Accounts or any Account which is controlled by other automated means". Users also agree "Not to use bots, scrapers, or automated scripts to access or interact with the Platform, or to circumvent moderation, pricing, or access controls". > > — Source: Fanvue Legal — General Terms & Conditions, read 2026-10-05 — https://legal.fanvue.com/terms-conditions > Updated 5 February 2026. "If there is any conflict between this Policy and the General Terms, the terms of this Policy shall prevail for the specific subject matter it governs." Permitted use includes "Chats: Listing chats, retrieving chat messages, creating new chats, sending messages, sending mass messages". > > — Source: Fanvue Legal — API Access & Usage Policy, read 2026-10-05 — https://legal.fanvue.com/api-policy No Fanvue page says when a reply bot turns your account into one that is "controlled by other automated means". So keep the account in your own hands: you log in, you read the chats, and you decide which messages a tool may send. If you want a firm answer for a fully automatic setup, ask Fanvue support before you build it. **What we recommend: tell your fans.** Fans pay for messages because they feel someone writes to them. If a bot or a chatter writes some of them, say so once in your bio and in the welcome message, for example: “Messages may be written with AI help or by my team.” This keeps you on the safe side of Fanvue's rules against misleading content and impersonation. It also protects you from chargebacks by fans who feel misled. ## Do you have to tell fans they are talking to an AI? Fanvue's general AI-content policy says you have to label AI-generated *media*. It does not cover conversations. [What AI OFM is](https://aiofmcity.com/blog/what-is-ai-ofm) explains that policy. For messages, the legal question is already worked through in [how money is made on Fanvue](https://aiofmcity.com/blog/where-fanvue-money-is-made#the-disclosure-question-answered-narrowly). The answer is clearer than for most EU AI Act questions: > Article 50(1) binds **providers**, not deployers: it requires whoever designs and supplies an AI system that interacts directly with people to build in a disclosure that they are an AI, unless that is already obvious from the circumstances. > > — Source: EU AI Act Art. 50(1), read 2026-09-09 — https://artificialintelligenceact.eu/article/50/ The important word is *providers*. A provider builds and supplies an AI system. A deployer only uses it. If you use Fanvue's own Reply, Rewrite or automated-message tools, Fanvue is the one who built the system and supplies it. In that case the rule in Article 50(1) applies to Fanvue. It does not apply to the creator who uses the tools. There is also the "obvious from context" exception. If your account is already labelled as an AI creator, you have a stronger argument that a fan should not be surprised when a reply comes from a system. ## If you build your own chatbot with the Fanvue API The chatbot from the API tutorial is a different case. Fanvue's documentation does not cover it, and no other page on this site does either. If you build a chatbot with Fanvue's API, your own code decides what to generate and when to send it, and it runs on your own schedule. You are no longer "using a system Fanvue supplies." You are building a new AI system and putting it into service. That is close to what Article 50(1) means by a *provider*. The scopes, the rate limit and the usage policy behind that API are in [Fanvue API and MCP server](https://aiofmcity.com/blog/fanvue-api-and-mcp). ### What the sources do not say: Finding guidance on whether building your own chatbot against a platform's API makes the operator the AI system's provider The first place to look for an answer was the European Commission's guidance on Article 50. This site already uses it for the editorial-responsibility exception and for the rule on deepfake media. It does not cover this case. Three sources were read: Fanvue's documentation, the Commission's guidance and the text of the Act. The Act only gives the general definitions of provider and deployer. None of the three says where "using a supplier's messaging feature" ends and "operating your own system built on their API" begins. One thing can be said. The more decisions your own code makes, and the fewer a Fanvue feature makes, the weaker the argument that you only use Fanvue's system. No one has drawn that line for this case yet. ## Fanvue's tool or your own bot: what to do 1. **If you use Fanvue's own Reply, Rewrite or automated-message tools**, Fanvue states no limit. The rule in Article 50(1) most likely applies to Fanvue, because Fanvue is the provider. 2. **If you build your own chatbot with Fanvue's API**, your position is very different. You are closer to being the provider of the system than a user of it, and Article 50(1) puts its rule on the provider. Fanvue's General Terms also forbid accounts "controlled by other automated means", so keep a person in charge of the account. 3. **In both cases, keep the AI label on your account up to date.** You need it for the media anyway. It also supports the "obvious from context" argument. 4. **Have a human check every message that a fan could read as a promise or a commitment.** No rule says you have to. But a mistake in this kind of message costs the most, whether the system is Fanvue's or your own. Automation matters this much because most of the money is made in messages. The numbers are in [how money is made on Fanvue](https://aiofmcity.com/blog/where-fanvue-money-is-made). The other labelling rules, besides Article 50(1), are explained in [what AI OFM is](https://aiofmcity.com/blog/what-is-ai-ofm) and, step by step, in [EU compliance for AI personas](https://aiofmcity.com/guides/eu-compliance-for-ai-personas). One step always stays with a real person, however much of the conversation you automate: [Fanvue verification and KYC](https://aiofmcity.com/blog/fanvue-verification-kyc). ## Frequently asked ### Is it against Fanvue's rules to fully automate replies to fans? Fanvue's documents point both ways. Its General Terms say you must not create "'bot' Accounts or any Account which is controlled by other automated means". Its API docs include a tutorial for an auto-reply chatbot that sends replies on a schedule, and its API policy allows sending messages through the API. No Fanvue page says when a reply bot makes your account one that is "controlled by other automated means". None of Fanvue's help pages on AI messaging states a rule to tell fans. ### Do I have to tell fans they are talking to an AI on Fanvue? Article 50(1) of the EU AI Act puts that rule on the provider, the one who builds and supplies the AI system. It does not put it on you for using the system. If you use Fanvue's own Reply or Rewrite tools, the rule most likely applies to Fanvue. If you build your own chatbot with Fanvue's API, you are closer to being the provider yourself, and the provider is the one the rule applies to. ### What is the difference between AI-assisted messaging and a chatbot on Fanvue? Fanvue's own "Reply" and "Rewrite" tools draft a message. A human reads it and sends it. Fanvue built that system and supplies it. The auto-reply tutorial in the API docs goes further. There, your own code decides what to generate and when to send it, on a schedule, and no human checks it. That is closer to running your own AI system than to using Fanvue's. ### Does Fanvue require personalising AI-drafted messages? No. Fanvue recommends it but does not require it. Its guidance says to add a fan's nickname or a note about something they mentioned to an AI draft, so the reply does not read as generic.