Policies

Enforcement Policy

Version 2.1 Last updated 2026-11-06

Purpose

This Enforcement Policy explains how Xive Inc. (“Xive”) responds when content, conduct, or accounts may violate our Terms of Service, Community Standards, monetization or streaming rules, or applicable law. It works together with our Moderation Model and Appeals Process—you should read those documents for how reports are triaged and how to challenge a decision.

  • We aim for consistent, explainable outcomes; not every case fits a template, and we may need time to investigate.
  • Nothing here limits our right to change features, refuse service where permitted by law, or cooperate with lawful requests from authorities.

How we detect and review issues

Enforcement usually starts with one or more signals that something may be wrong. We combine automation with human judgment.

  • User reports and safety workflows submitted in product or through designated channels.
  • Automated detection, including classifiers, spam or abuse heuristics, and integrity systems designed to reduce harm at scale.
  • Trusted flagger or partner programs where we operate them and where law or contracts allow.
  • Notices from payment, app store, and integration partners, such as Stripe, Apple, Google, Twitch, Kick, and Discord, that an account or its content breaks their rules.
  • Human review for escalations, edge cases, appeals, and categories that automated systems are not allowed to decide alone.

Depending on severity and product design, automated signals may trigger warnings, queue content for human review, reduce distribution, or apply temporary limits while a decision is pending. We align those steps with law, user safety, and transparency expectations.

Types of enforcement

We may apply one or more of the following, alone or in sequence. The list is illustrative, not exhaustive.

  • Content-level: remove or restrict a specific post, clip, message, thumbnail, stream segment, or other item; add age restrictions or labels; geo-block or withhold distribution where the law requires.
  • Broadcast-level: end a live stream while it is running, and withhold or delete its recording. Ending a broadcast also rotates your stream key, so you will need to copy the new one from Go Live before you can stream again. This is not a streaming ban, which is a separate account-level action.
  • Feature-level: temporary limits on posting, streaming, messaging, monetization, discovery, or other capabilities while we investigate or as a proportionate sanction.
  • Account-level: warnings; time-bound suspensions; permanent termination of access; withholding or reversing payouts when we identify fraud, chargeback abuse, or serious policy breaches, subject to law and payment-partner rules.

Warnings, standing, and automatic chat enforcement

Warnings build up into an account standing, so repeated problems are handled more firmly than a single mistake:

  • Warnings count toward your account standing for a set period, currently 90 days. Three active warnings put your account in “At Risk” standing, where further violations are more likely to lead to suspension. When a temporary suspension ends, your account returns in At Risk standing. You can see your standing and any active warnings in Settings → Account Standing.
  • Serious violations can lead to suspension or termination without earlier warnings, as described above.
  • Copyright strikes are counted separately from the standing above, on their own clock: a strike counts for 12 months, and three in 12 months can end your account under the repeat-infringer policy in our Terms. You can see and appeal them in the same place. Not every copyright removal carries a strike — where we remove something on our own initiative, without a rights holder claiming it, we record a strike only where we are satisfied the material was infringing.

Live chat is also moderated automatically, without a person reviewing each decision:

  • A filter blocks chat messages that break our Community Standards, such as slurs, threats, scams, and spam. The sender is told the message was not posted.
  • Serious blocked messages, such as slurs, threats, scams, and doxxing, also count as a strike in that stream. Repeated strikes lead to automatic timeouts from that chat that get longer each time, up to 7 days.
  • We never ban anyone from a chat automatically; only the creator or their moderators can. They can also lift an automatic timeout at any time. Strikes expire over time.

Age and identity checks

Some places require us to confirm that account holders are 18 or over, and we offer a free identity check and, in some places, a selfie age check for that. How we respond when someone tries to get around those checks depends on what we can establish:

  • Signals that do not add up — for example an account seen in a place that requires a check, or whose profile location names one, that then tries to avoid it — lead us to ask for the identity check. Until it is passed, the account can only read our policies, download its data, or delete itself. This is a safety step, not a penalty.
  • Deliberate deception — a borrowed, fake, or altered ID, a spoofed or stand-in selfie, or new accounts made to dodge a check — can lead to suspension or permanent termination, and to action on linked accounts.
  • Evidence that an account holder is under 18 leads us to close the account under our minimum age rule, however we learn of it.

Using a VPN, proxy, or similar service is not a violation on its own. It does not remove an age check: once we have seen an account in a place that requires one, the requirement stays until the check is passed or our team clears it. You can appeal any of these decisions through our Appeals Process.

How we weigh severity and context

The same rule text can call for different responses depending on facts. Reviewers consider factors such as:

  • Nature and seriousness of the conduct (harm, illegality, deception, or risk to minors or vulnerable people).
  • Apparent intent, repetition, and history of prior violations or attempts to evade enforcement.
  • Scale of harm or risk to individuals, groups, or the integrity of the platform.
  • Public interest, documentation, and newsworthiness where those concepts apply under our policies.
  • Regional legal requirements, court orders, or regulator guidance that affect what we must do or must not do.

Certain single incidents—such as credible threats, child sexual exploitation, terrorism promotion, malware distribution, or similarly severe harm—may result in immediate termination or referral to law enforcement where appropriate, without a graduated warning sequence.

We publish or provide information about our enforcement practices where law, contracts, or industry norms require it. We also balance transparency with:

  • Active investigations where premature disclosure could compromise evidence or safety.
  • Security-sensitive details that would help bad actors bypass controls.
  • Privacy obligations toward reporters, victims, and third parties who did not choose to be named in a public case summary.

When we must notify users, rights holders, or authorities under statute or court order, we do so in line with those instruments and with advice from legal counsel when needed.

For how to appeal a strike or account action, use the Appeals Process policy and any in-product appeal links we provide with the enforcement notice. For how queues and reviewers are organized, see the Moderation Model. For baseline rules of conduct, see the Community Standards and Terms of Service.

Contact

After you have read the Appeals Process, you may contact [email protected] for questions about enforcement that are not resolved by self-service or appeal tooling.

More policies