Safety Center
How SkillFans reduces risk around adult-only access, AI transparency, private media delivery, reports, moderation, and abuse prevention.
1. Safety principles
SkillFans is designed as an adult-only AI virtual creator product. The safety model focuses on age restrictions, AI transparency, rights and consent controls, private media delivery, review queues, report handling, role-based admin access, audit logs, and abuse prevention.
No safety system is perfect. SkillFans uses a layered model to reduce risk, respond to reports, and preserve operational visibility rather than relying on a single control.
2. Adult-only access
SkillFans is only for adults 18 and older. Anonymous users must confirm the age gate before browsing. Logged-in users may have age-gate confirmation recorded to their profile so operators can verify that the account passed the adult-only gate.
Users who are under 18, who are not legally permitted to access adult-oriented online services, or who are trying to help a minor access SkillFans must not use the service.
3. AI transparency
SkillFans V0 requires AI virtual creator labeling. Creator cards, creator detail pages, asset metadata, upload workflows, and policy surfaces should not mislead users into believing synthetic creators are real people.
Missing or misleading AI labeling can be reported and may lead to changes requested, asset rejection, creator pause, or removal.
4. Minor safety
SkillFans does not allow minors to use the service and does not allow content that depicts, simulates, describes, requests, or sexualizes minors or age-ambiguous subjects. Youth-coded, school-age, childlike, teen-coded, or otherwise underage-presenting sexual content is prohibited.
Reports involving suspected child sexual exploitation, grooming, sextortion, or minor safety may lead to immediate restriction, evidence preservation, and escalation to appropriate safety organizations or authorities where required or appropriate.
5. Consent, likeness, and rights safety
SkillFans prohibits non-consensual sexual content, revenge sexual content, hidden-camera content, leaked private media, sexual deepfakes of real people without consent, unauthorized use of a real person's likeness, and rights-infringing uploads.
Reviewers may request rights-holder details, AI workflow context, provenance information, or consent evidence before approving or restoring content.
6. Private media delivery
Sensitive media is intended to be stored in private object storage rather than public buckets. Public pages receive approved metadata and safe preview information. Eligible file access is handled through permission checks and short-lived signed URLs.
Raw storage references are not intended to be exposed to ordinary browsers. If SkillFans detects unauthorized exposure, scraping, sharing, or bypass attempts, it may revoke access, rotate storage paths, block delivery, or take enforcement action.
7. Review before publishing
Creator applications and uploaded assets may be reviewed before publication. Reviewers may approve, reject, request changes, pause, ban, restore, reopen, or remove content based on policy, rights, consent, AI-labeling, safety, provider, or legal concerns.
Approved content remains subject to later review if new reports, legal notices, provider signals, or safety evidence appear.
8. Reporting controls
Users can report creator or asset concerns for minor safety, non-consent, rights issues, AI-labeling, spam, or other reasons. Reports are designed to create a case record with target type, target ID, reason, details, status, and operator review history.
Admins, reviewers, or risk-ops users may mark cases for manual review, resolve them, dismiss them, or take related content and account actions.
9. Vote, invite, and Growth Point integrity
SkillFans uses server-side daily vote limits, Growth Point ledger records, task claim controls, invite-code status controls, referral activation rules, risk events, and audit logs to protect launch metrics from manipulation.
Fake engagement, automated voting, point farming, invite abuse, referral fraud, and attempts to alter ledgers or unlock records may lead to account restrictions, GP adjustments, invite freezes, report escalation, or bans.
10. Admin access and auditability
Admin and moderation surfaces are separated from ordinary public browsing and are protected by Supabase roles and server-side checks. Ordinary users should not be able to enter admin consoles or call admin mutation RPCs.
Admin actions are intended to create audit records so that creator approvals, asset decisions, risk updates, invite freezes, rule changes, and other operational actions remain reviewable.
11. What users should do
- Keep your email, OAuth account, devices, browser sessions, and magic links secure.
- Do not share locked content, signed URLs, raw storage references, or private creator assets outside authorized product flows.
- Report suspicious creators, assets, AI-labeling issues, rights issues, consent issues, or minor-safety concerns from the relevant page whenever possible.
- Avoid submitting unnecessary personal information in report details unless it is needed to explain the issue.
- Call local emergency services first if someone is in immediate danger.
12. System limits
SkillFans cannot guarantee that every harmful upload, false report, impersonation attempt, rights dispute, storage misuse, or account abuse attempt will be detected immediately. Safety controls reduce risk, but users, creators, reviewers, rights holders, and service providers all play a role in identifying and resolving problems.
13. Safety contact
For legal, copyright, privacy, consent, minor-safety, or urgent platform-safety concerns, contact legal@skillfans.org. Use the in-product Report button when available so the case includes target IDs and can enter the moderation queue directly.