← Cold XVI

The Cold XVI Content Standard

What we measure, what we celebrate, and the one line we will not cross. We govern harm, not subject matter — and this explains exactly what that means and how it is enforced.

Last updated: 27 August 2026

A standard for what we measure, what we celebrate, and the one line we will not cross.

Why this document is written the way it is

Most content policies are decency policies. They ask a single question — does this depict something bad? — and remove whatever does. We do not run a decency policy, and we want to tell you plainly why, because the reasoning is the policy.

A decency standard cannot tell the difference between a work that enacts an evil and a work that anatomizes one. It bans the second along with the first, because all it can see is subject matter. By that standard the greatest books ever written are contraband, and the standard is not strict — it is simply illiterate. Great art has always gone to the darkest places precisely in order to say something true about them. A rule that flinches at darkness cannot host art at all. It can only host the inoffensive, which is a different and much smaller thing.

So we draw our line somewhere else. We govern harm, not subject matter. A verse can be as violent, profane, transgressive, or morally black as the art requires, and we will measure it without flinching and, where it earns it, celebrate it. What we will not do is participate in a real harm to a real person. That is the whole of the line, and everything below is an unfolding of it.

We do not claim that everything permitted here is good. Some of it is not. We claim something narrower and more honest: that the only line a platform can draw competently, consistently, and without arbitrariness is the harm line — and that a platform which pretends to have found a workable standard of decency is lying, first to its users and eventually to itself.

Two acts, two standards

Cold XVI does two different things with a verse, and they are governed differently because they are not the same moral act.

We score. The engine measures craft — pocket, rhyme density, syncopation, drum lock, delivery. This is a judgment about the thing as made, and it is indifferent to whether what the verse depicts is good. Scoring is not endorsement any more than a scansion tool endorses the Inferno by scanning it. At this layer there is no such thing as content too dark to measure. An engine that refuses to read the blackest bars is a broken instrument with no reason to exist. We measure everything.

We amplify. When the platform ranks a verse, features it, builds a post around it, it does something more than measure — it celebrates. Celebration is a different act with a different weight, and it is here, and only here, that the harm line does its work. The question at this layer is never is this verse dark? It is does our lifting this verse up participate in a real harm to a real person? Almost always the answer is no, and the verse flows. Sometimes the answer is yes, and it does not.

Hold these two apart and everything else is clear. Collapse them and nothing is.

The line: harm to a real person

We do not host, score, or amplify content that inflicts a real harm on a real, identifiable person. In practice this means:

  • Derision of real victims. Verses that name and mock actual, identifiable people who were killed or assaulted — where the “art” is functionally a desecration of a real person whose family is alive to read it. The harm here has a real object. This is not us objecting to violence in the abstract; depicted, fictional, and generalized violence is fully protected. It is us refusing to make the platform a co-participant in the derision of someone real.
  • Doxxing and targeted exposure. Content that functions to expose a real person’s private information or location, whatever formal shape it takes.
  • Credible threats and incitement to live harm. Content that is a genuine threat to an identifiable target, or that meaningfully incites real, ongoing violence against real people — as distinct from the vast body of art that depicts violence.
  • Harassment of someone who isn’t in the ring. Beef is the art form and we don’t referee it. Come for another user’s bars, their persona, their catalog, their whole reputation as a rapper — that’s the tradition and it’s welcome here. What isn’t welcome: pursuing someone who hasn’t engaged, continuing after they’ve blocked or declined, or leaving the craft entirely to go at a person’s real identity, family, job, location, or private life. The test is whether it’s a bar aimed at a rapper or a campaign aimed at a person.

Notice what is not on this list: darkness, profanity, drugs, transgression, braggadocio, moral blackness, or any depicted subject matter whatsoever. None of it is objectionable here, because none of it is a harm to a real person. That is the point.

Nor is competitive disrespect. Rap is a combat art, and the diss is one of its oldest and best forms. Two rappers going at each other is not harassment — it is the genre working correctly, and a platform for the craft that couldn’t host it would be missing the craft.

These four categories are the whole list, and they are the same four you will see in the app. When you report something, the reasons you are offered are these — not a longer list of things we find distasteful. We report against the standard we actually hold.

The floor beneath the line

There is a category we exclude that is not reached by our harm principle at all — it is reached by the law, and by the simple fact that some things are not depiction in the first place.

Child sexual abuse material is the clearest case. It is prohibited absolutely, with no craft exception, no context exception, and no exception of any kind. The reason is not that it depicts something objectionable — that is a decency argument, and we have already refused decency arguments. The reason is that it is not a depiction at all. It is the documentary record of a real crime against a real child. It is direct harm to a real person in the most literal sense the words allow. The distinction that governs our entire standard — depiction versus record-of-real-harm — draws this line for us from the inside, on exactly the grounds we already hold.

Alongside this, we operate within the law. Certain content is prohibited on this platform because the legal order we ship under prohibits it, independent of what our own harm principle would say. We are candid that this floor sits below our ethics and that we are bound by it regardless. A platform is not a novel; it does not get to be Nabokov. It invites, hosts, scores, and amplifies the work of thousands of strangers, and it does so inside a legal system, a set of platform rules, and a payment infrastructure, all of which bind it. We will not pretend otherwise, and we will not gamble the existence of this platform — and everyone’s work on it — on the difference between what we believe and what we are permitted to operate.

For a platform devoted to the craft of rap, this floor costs the art nothing. The verses we exist to measure are dark about things; they are never the record of a real harm. The line and the art do not collide.

How this gets enforced

Principles that cannot be operated are decoration. Ours run on the following mechanisms, present throughout the app:

  • Report. Every piece of content carries a report action. Anything a user believes crosses the harm line can be flagged in one step.
  • Block. Every user profile carries a block action. You can remove any other user from your experience entirely. For conflict between users, this is the primary remedy and we expect it to resolve most of it — moderation is the backstop, not the first move.
  • Screen. Submitted content is screened automatically. We want to be exact about what that does, because most platforms are vague here and the vagueness is doing work for them.
  • Review and response. Reports enter a moderation queue and are actioned within 24 hours. Content found to cross the line is removed; users who repeatedly cross it lose access.
  • Notice and appeal. If we remove your content or restrict your account, we tell you — what was removed, and which part of this standard it was found to cross. You can contest that decision, and a person reads what you write.
  • Contact. Write to safety@cold16.com about anything on this page. It is read.
  • Age assurance. This is an explicit-content platform. Content is rated accordingly and gated by age. The Service requires users to be 16 or older; the Label Discovery Program requires 18 or older.

What the screen actually does

Content you submit goes live immediately. It is not held in a queue waiting for a human to approve it, and your position on a leaderboard is never waiting on a moderator — on a competition that runs for a day, that would be the same as not entering at all.

The automated screen runs alongside that. It does not decide anything. It reads what you submitted and, where it believes it has found something that crosses the harm line, it quotes the exact words and names the real person it believes is being harmed, and moves that submission to the top of a human’s queue. A person makes every decision. This is deliberate: an automated system that could remove your work on its own, with no human judgment in the path, is not something we are willing to run.

The screen is also built to a specific and narrow question. It is never asked whether a verse is offensive, dark, violent, or profane — all of which are protected here, and all of which describe most good rap. It is asked whether a real, identifiable person is the object of a harm. A finding that cannot name that person and quote the words is discarded automatically, before anyone sees it.

There is one exception, and we would rather say it than have you discover it. For the absolute floor — child sexual abuse material, and content that is flatly unlawful — detection hides the content immediately, before human review. Everything else stays up while a person looks at it. We are not willing to write “everything posts optimistically” and quietly mean “including that.”

This document governs harm. It does not govern infringement, which is a separate question with a separate procedure and its own legal clock.

If you believe your work has been used here without permission, that is handled under the Copyright and Repeat Infringer Policy, which explains how to send a notice, how to counter one, and what happens in between. We terminate the accounts of repeat infringers, and that policy sets out what that means in practice.

Reporting something as “not their verse” — a pro acapella, or somebody else’s bars — is different again, and it lives in the app. It is the anti-cheat path on a platform whose entire pitch is honest numbers, and it is the report we most want you to file.

Where we hold ourselves

We hold ourselves to this standard — the harm line as written here — not to a vaguer or broader one. We have deliberately drawn a narrow line so that we can actually enforce it. A platform that promises to remove everything “offensive” has promised something it cannot keep and handed a weapon to anyone acting in bad faith. We promise something smaller and truer, and we keep it.

User Agreement (Content Clauses)

By creating, submitting, or scoring content on Cold XVI, you agree to the following. Cold XVI is operated by Toggle Media, LLC, a Florida limited liability company, and these clauses form part of the Cold XVI Terms of Service and End User License Agreement, which is governed by the laws of the State of Florida.

  1. No objectionable content, defined by harm. You will not submit content that inflicts a real harm on a real, identifiable person, including: content that names and derides real victims of violence; content that exposes a real person’s private information or location; credible threats against, or incitement to real violence toward, identifiable people; and harassment of a user who has not engaged you, including continuing to target a user after they have blocked or declined, or attacking a user’s real identity, family, employment, location, or private life. Competitive disrespect directed at another user’s bars, persona, or standing as a rapper is not a violation of this clause.

  2. Absolute prohibitions. You will not, under any circumstances, submit content that constitutes child sexual abuse material, that is otherwise unlawful under applicable law, or that the platform is prohibited from hosting by its operating requirements. These prohibitions admit no exception.

  3. No tolerance for abusive users. Conduct that crosses from competition into harassment, as defined in clause 1 and in this Content Standard, is prohibited. We reserve the right to remove content and terminate accounts that violate these terms.

  4. Craft scoring is not endorsement. Scores reflect a measurement of craft and do not constitute an endorsement, by Toggle Media, of any conduct depicted in the scored content.

  5. Reporting and moderation. Content may be reported through in-app tools. We review reports and act on violations within 24 hours, including removal of content and, for repeated or egregious violations, termination of access.

  6. Copyright. Claims of infringement are governed by the Copyright and Repeat Infringer Policy, and accounts of repeat infringers are terminated.

  7. Age. This platform contains explicit content. You must be 16 or older to use the Service, and 18 or older to participate in the Label Discovery Program. You consent to age assurance where applicable.

  8. Enforcement. We may remove content, and suspend or terminate accounts, that violate these terms, this Content Standard, or applicable law.

This document states a standard and a philosophy. It is not legal advice. The User Agreement clauses above form part of the Terms of Service; where the two differ, the Terms govern.