Moderation and Reporting Policy
AI72 moderates to protect a safer, respectful, consent-based controlled introduction environment, not to punish users for its own sake. This policy explains how reports, blocks, safety signals, AI-assisted tools, human/admin review, evidence retention, and appeals may be handled.
1. Purpose
AI72 may moderate to protect platform safety, reduce abuse, enforce the Terms and Community Guidelines, respond to user reports, address illegal content and safety risks, and support support/legal processes where needed.
2. Moderation philosophy
AI72 may use user reports, block records, platform rules, safety signals, automated tools, AI-assisted tools, service providers, and human/admin review. No system is perfect; harmful content may be missed, safe content may be flagged, and context may be misunderstood.
- Moderation may depend on severity, available evidence, user history, repeat behavior, legal obligations, and user safety.
- A report does not automatically mean that a violation was found or that enforcement will occur.
- More urgent or severe risks may lead to faster or stronger action.
3. What can be reported?
Where available, users may report profiles, profile text, profile photos, profile media, temporary intro sessions, permanent chats, text messages, voice messages, audio, photos, videos, Shared Media, video-call behavior, verification fraud, fake-camera/replay/spoofing, spam, scams, money requests, harassment, threats, sexual content, minor risk, fake profiles, impersonation, deepfakes, platform abuse, and technical abuse.
4. Who can report?
Users may report violations they encounter in temporary intros, permanent chats, profile views, media views, Shared Media, support, or safety flows. AI output is reportable only where an AI output surface is actually available and reportable. AI72 may also receive safety signals from blocks, support tickets, admin flags, service-provider notices, platform notices, or valid legal requests.
5. Blocking
Blocking is a boundary and safety control. Depending on product context, blocking may stop new messages, prevent future temporary intro matching, stop or limit permanent chat interaction, reduce profile visibility, and create safety signals for review.
- Blocking does not automatically delete historical records, reports, messages, media, audit logs, or legal/safety evidence.
- Some records may be retained for safety, dispute, legal, audit, or abuse-prevention purposes.
- Blocking is not emergency intervention; contact local emergency services or law enforcement for urgent danger.
6. Moderation pipeline
A moderation workflow may begin with a report, block, safety signal, support ticket, admin flag, or legal request. It may then involve safety queue intake, triage, AI-assisted support where used, human/admin review where appropriate, decision, user notice where appropriate, appeal/support review, and evidence retention.
- Not every report goes through every step.
- Child safety, exploitation, violence, legal requests, or severe abuse may be escalated faster.
- No-action is also a possible moderation outcome.
7. AI-assisted moderation
AI or automated systems may assist with prioritization, image/media safety review, profile media moderation, translation/transcription support, abuse-pattern signals, or report-review context where used. AI is assistive; final outcomes are not always made by AI.
- AI may miss harmful content or flag safe content incorrectly.
- AI may misunderstand context, language, humor, culture, consent, or intent.
- AI72 does not claim that all chats are automatically moderated by AI; reports and selected media/safety workflows may be reviewed with automated tools and human review.
8. Human/admin review
Human or admin review may be used for minors, CSAM, grooming, sexual exploitation, scams, violence, threats, sextortion, stalking, harassment, fake profiles, impersonation, verification fraud, media review, appeals, and legal requests.
- Human review may be limited by available evidence, privacy, security, anti-abuse, and applicable law.
- Unnecessary or casual access by unauthorized people is not intended.
9. Evidence and retention
When a report, block, moderation action, support request, safety incident, or legal request occurs, AI72 may preserve relevant evidence such as report text/category/timestamps, account ids, profile fields/media, temporary/permanent chat metadata, message text, voice/audio/transcript/translation/TTS metadata, media access state, storage paths, verification status/signals, block records, prior enforcement, appeal history, support communications, admin notes, audit/device/security/diagnostic logs.
- Evidence may be retained to investigate reports, protect users, prevent repeat abuse, review appeals, handle disputes/legal processes, and keep audit history.
- Evidence should not be accessed unnecessarily; access should be limited to authorized systems/personnel with legitimate safety, support, legal, security, or compliance need.
- Retention depends on the Privacy Policy, applicable law, deletion requests, severity, appeal status, and abuse-prevention needs.
10. Media moderation
Profile photos, profile media, temporary intro media, permanent chat media, Shared Media, voice, video, and related metadata may be reviewed when reported, flagged, public/profile-visible, or part of a safety workflow.
- Media may be marked pending, allowed, hidden, disabled, quarantined, restricted, expired, removed, or retained for review.
- Hidden, deleted, expired, report-quarantined, disabled, restricted, or removed media may not appear in Shared Media or user-facing views.
- Media access may depend on session state, product rules, signed access, moderation state, legal obligations, and safety rules.
11. Repeat offender policy
AI72 may consider repeated reports, repeated blocks, prior warnings, prior restrictions, ban-evasion attempts, linked-account signals, severity, pattern, and context.
- Actions may include notice/warning, content removal, media disablement, visibility restriction, feature restriction, temporary suspension, permanent ban, re-verification requirement, or future account prevention.
- Severe child-safety, CSAM, exploitation, credible threats, trafficking, non-consensual intimate media, sextortion, fraud, or legal orders may lead to stronger action without earlier steps.
12. False reports
Reports must be made in good faith. Intentionally false reports, revenge reports, spam reports, coordinated reporting, attempts to silence another user, false child-safety accusations, forged evidence, or moderation-system manipulation may lead to enforcement.
13. Child safety
AI72 is only for users aged 18 or older. CSAM, grooming, minor sexualization, live exploitation, trafficking, seeking romantic/sexual contact with minors, or behavior creating child-safety risk is zero tolerance.
- For suspected minors, CSAM, grooming, trafficking, or child-safety risk, AI72 may preserve evidence, restrict features, suspend/terminate accounts, and share with competent authorities/reporting bodies where legally required or appropriate.
- Users should report suspected minors or child-safety risks.
14. Violence, extortion, sexual abuse, hate, and illegal activity
Threats, extortion, blackmail, revenge-porn threats, sextortion, stalking, harassment, hate, extremist activity, terrorism, illegal goods/services, trafficking, sexual exploitation, non-consensual intimate media, doxxing, privacy invasion, malware, phishing, fraud, and similar behavior may lead to enforcement. In urgent physical danger, contact local emergency services or law enforcement; AI72 is not an emergency service.
15. Off-platform conduct
Serious off-platform harassment, blackmail, threats, fraud, stalking, non-consensual media sharing, child-safety risk, or violence may lead to moderation if it affects AI72 users, platform safety, account integrity, or legal obligations.
16. Appeals
If you believe a moderation, content, media, feature, or account decision was wrong, you may appeal through support or any available in-app appeal path. Appeals may involve human/admin review.
- An appeal does not mean that content will be restored, a feature will be unlocked, or an account will be reopened.
- AI72 may limit details for safety, privacy, anti-abuse, or legal reasons.
17. Transparency and limited explanations
AI72 may provide notices or explanations where appropriate, but some decisions cannot be fully explained because of reporter safety, child safety, privacy, anti-abuse, fraud prevention, security, legal obligations, or law-enforcement sensitivity.
18. Law enforcement and legal requests
AI72 may preserve, disclose, or provide relevant records when required or permitted by applicable law, valid legal process, emergency disclosure rules, child-safety obligations, or rights/safety protection. AI72 does not provide data merely because someone asks; requests may be reviewed for validity, authority, scope, and legal basis.
19. What AI72 does not guarantee
AI72 does not guarantee it will detect every scam, fake profile, underage user, abusive user, harmful message, unsafe media, harassment pattern, or legal violation. Not every report is accurate, not every moderation decision is perfect, not every appeal succeeds, and not every harmful user is removed immediately.
20. Linked documents
This policy should be read with the Terms of Service, Privacy Policy, Community Guidelines, Safety Center, AI and Automated Processing Disclosure, Account and Data Deletion, Notifications and Chat Actions, Data Safety and Permissions, and Support documents.