Overview
Velaris Management Group LLC (“Xive”) moderates content and accounts to enforce our Community Standards, Terms of Service, monetization and streaming rules where they apply, and applicable law. The goal is to protect people on the service, preserve trust in the platform, and keep enforcement predictable enough that creators and viewers know what to expect.
- Moderation blends trained human reviewers with automated systems; both follow the same published policies and are updated when those policies change.
- Neither humans nor machines are perfect; we use quality controls, sampling, and training to reduce inconsistency and mistaken takedowns.
- Final enforcement actions (warnings, restrictions, removals) are described in our Enforcement Policy; if you disagree with an outcome, use the Appeals Process.
Queues and prioritization
User reports, trusted-flagger inputs, integrity signals, and automated classifiers feed prioritized work queues. Not everything is reviewed in real time; we allocate capacity to the highest-risk items first.
- Accelerated paths—where staffing and law allow—for credible threats, child safety, self-harm risk, terrorism or violent extremism, and similarly severe categories.
- Standard queues for spam, harassment, nudity or sexual content disputes, IP reports, fraud, and general Community Standards violations.
- Backlogs during incidents or viral events may delay lower-severity items; we do not use that delay as a reason to ignore imminent harm when we become aware of it.
Human review
People remain responsible for context that models handle poorly and for decisions that carry high stakes for expression or safety.
- Trained moderators interpret policy against the full context: satire, news, art, regional norms, and mixed-language or cultural cues.
- Humans decide or confirm many escalations, edge cases, repeat-offender patterns, and appeals that automation cannot lawfully or safely close alone.
- Reviewers record outcomes in internal tools where our product supports it, so we can audit for consistency, coach teams, and improve policy wording.
Automation
Automated systems scale detection and friction—never replacing the obligation to treat users fairly under our policies and the law.
- Typical uses include spam and malware filtering, illegal-content hashing workflows where deployed, rate limits, risk scoring, and triage into human queues.
- Automated blocks, filters, or distribution limits may be reversed or narrowed after human review or a successful appeal.
- We test models and rules for false positives and disparate impact where feasible, and we iterate when error patterns show up in appeals or audits.
Product surface and regional context
The same Community Standard may be applied differently depending on where content appears and where users live, when the law or product design requires it.
- Live streaming, messaging, clips, profiles, and commerce flows can each have extra rules or technical controls described in their respective policies or in-product disclosures.
- Some regions require geo-blocking, age gating, default private settings, or availability limits instead of—or in addition to—global removal.
- Labels, interstitials, or reduced distribution may be used when removal is not required but viewers should receive a warning or extra context.
Appeals
If you believe a moderation decision was wrong, follow the Appeals Process policy and use any in-product appeal link tied to the notice you received. That process coordinates with the Enforcement Policy for account-level outcomes.
Contact
General moderation and safety questions: [email protected]. For appeals, use the Appeals Process so your request is tracked with the right references.