Draft: this document is not final; items marked “to be filled” will be completed before launch.

Content Policy (Draft)

Version 2026-10-24 · Draft, not yet finalised by counsel. Items marked "to be filled" will be completed by the operator before launch.

The idol images, chat replies and generated images / videos / music on this site are all AI-generated.

1. Scope

This policy covers everything you put in and get out of the Service: chat messages, avatar and appearance descriptions, stage prompts, lyrics, uploaded reference images, generated works, the shared page and external publishing.

2. Content Boundaries

1. Safe for work (SFW) only: no sexual, suggestive or adult content, no nudity and no sexualised depiction.
2. No characters who are minors: do not create, describe or refer to a character who is under 18, looks under 18 or could be read as under 18, and never place any minor in a romantic or suggestive situation. Fictional settings, artistic intent and drawing style are not exceptions.
3. No real identity data: do not enter or generate other people's ID numbers, passport numbers, bank card numbers, social security numbers or similar real identity data, and do not publish other people's contact details or addresses.
4. Respect real people and third-party rights: without permission, do not use a real person's likeness, voice or name, do not impersonate a real person or artist, and do not upload songs, lyrics or images you have no right to use.
5. No unlawful use or abuse: including fraud, incitement to crime, mass attempts to get around content checks, and attacks on the Service.

3. How We Enforce It

• Input check: stage prompts, lyrics, chat messages that would start a generation and avatar descriptions go through a multilingual keyword and rule check before any points are deducted. On a match, the request is refused and you are told which category it broke; no points are deducted and nothing is sent to any generation service.
• Final prompt check: the final prompt rewritten by AI is checked again before it goes to a generation service, and the operator can switch on an additional AI review. When it is rejected, the generation is not submitted and the points deducted for it are returned.
• Default rules for AI roles: the default system prompts of AI roles such as chat, director, stylist and producer include the SFW and no-minor-characters rules.
• The generation services' own moderation: some generation services refuse or filter content; when they return a moderation signal, we record it.
• Manual takedown: after an administrator confirms a violation, they take the work down:
  • the work is marked as taken down in the database and no longer appears in the shared page or gallery responses;
  • the system tries to delete the file in object storage and the local copy on the server; a failed deletion is recorded and retried manually by an administrator;
  • every takedown leaves an audit record.
• Limits of a takedown: copies that visitors downloaded or screenshotted, or that browsers or CDNs cached (media files carry a cache setting of up to one year), are outside our control.
• Account action: for repeated or serious violations we may suspend new generations or ban the account.

4. Reports and Complaints

• Contact for reports and complaints: to be filled.
• Process: receive → verify → take down → reply → record.
• Reply time: to be filled.

5. Version and Language

• This policy is available in Chinese, English and Malay. If the versions differ, the Chinese or English version prevails.