Community Guidelines
Effective August 21, 2026 · Version 2026-08-21.1
The rules for what you may post, publish, and send on NovelVerse, what happens when a rule is broken, and how to report something or challenge a decision. These Guidelines elaborate the Terms of Use; the Terms remain the binding agreement.
On this page
- How to read these Guidelines
- Safety of minors
- Violence, threats, and self-harm
- Harassment, hate, and privacy
- Authenticity and platform integrity
- Rights, attribution, and AI assistance
- Mature content and labelling
- What happens when a rule is broken
- Reporting something
- Challenging a decision
- How these Guidelines change
How to read these Guidelines
NovelVerse is a place for fiction, and fiction goes to difficult places. These Guidelines are written to protect people, not to protect readers from subject matter. A novel may depict cruelty, crime, addiction, abuse, or despair without breaking a single rule below — depiction is not endorsement, and a story about a terrible thing is not the terrible thing.
What the rules restrict is conduct: material that hurts real people, that sexualises children, that deceives, or that attacks the platform others depend on. Where a rule touches creative work, it says so explicitly and states where the line falls.
Every rule carries a permanent identifier. When a moderation decision affects you, the notice you receive names the exact rule and the version of these Guidelines in force at the time, so you can read the same text the decision was made against.
Safety of minors
These are the rules with no exceptions, no artistic defence, and no appeal to intent. They apply to published chapters, private drafts, comments, messages, profiles, and uploaded images alike.
CG-MINORS-01 — Child sexual abuse material is prohibited absolutely, in every form, including text and generated images.
This covers photographic material, illustration, generated imagery, and written text. There is no fictional exemption: a written scene depicting the sexual abuse of a child is covered by this rule exactly as an image would be.
NovelVerse preserves and reports this material as Republic Act 11930 requires. Enforcement is immediate and does not wait for an appeal.
Possible outcomes after human review: remove the content, ban the account.
CG-MINORS-02 — Characters presented as children may not be sexualised, whatever their stated age.
A character's declared age does not control this rule. What controls it is how the character is presented — body, role, schooling, dependence, and the way the text frames them. A character written as a child and labelled an adult is a child for the purpose of this rule.
Non-sexual depiction of children in fiction is entirely permitted, including children in danger, in grief, or in harm's way. The prohibition is on sexualisation.
Possible outcomes after human review: remove the content, ban the account.
- Not permitted: a sexual scene involving a character described with a child's body and circumstances, with a line asserting they are 500 years old.
- Permitted: a novel in which a child is abused, written so the abuse is depicted as harm rather than as erotic material.
CG-MINORS-03 — NovelVerse accounts are for adults; you must be 18 or older.
You confirm your age when you create an account. Accounts found to belong to someone under 18 are closed, and material suggesting an account holder is a minor is treated as a safety matter rather than a rules technicality.
Do not solicit personal contact, images, or a relationship from someone you believe to be a minor. Report it instead.
Possible outcomes after human review: restrict the account or feature, ban the account.
Violence, threats, and self-harm
The distinction throughout this section is between a story and an act. Writing violence is a story. Directing violence at a real person, or supplying a real reader with the means to hurt themselves, is an act.
CG-HARM-01 — Do not threaten anyone, or make a real person the target of violent intent.
This covers direct threats, conditional threats, veiled threats, and encouraging others to carry one out. It applies to comments, messages, profiles, and published text where a real, identifiable person is the target.
A threat is judged by what a reasonable reader would take it to mean, not by whether the author says they were joking.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-HARM-02 — Do not incite or celebrate violence against real people or groups.
Calls for violence, praise for an attack, and material designed to recruit for one are prohibited regardless of the cause.
Fiction depicting atrocity — including from inside the perspective of those committing it — is not covered by this rule. What is covered is text that functions as advocacy addressed to the reader.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-HARM-03 — Do not encourage self-harm or supply methods, and never as advice to a real reader.
Instructional detail about suicide or self-injury, encouragement addressed to a person, and material framing self-harm as a solution are prohibited.
Fiction may depict suicide, self-injury, and the despair around them. Depicting a character's method is not the same as instructing a reader in one; where a passage reads as instruction rather than narrative, it falls under this rule.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature.
CG-HARM-04 — Do not post real graphic violence or death imagery as shock material.
Photographic or video depiction of real injury, death, or mutilation posted for its effect, without documentary or narrative purpose, is removed.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature.
Harassment, hate, and privacy
Criticism is not harassment. A harsh review, an unfavourable comparison, and a public disagreement about someone's writing are all ordinary here. What these rules restrict is conduct aimed at a person rather than at their work.
CG-CONDUCT-01 — Do not target a person with sustained or degrading abuse.
Repeated unwanted contact, sexual harassment, degrading remarks about a person rather than their work, and following someone across surfaces to continue an argument are all covered.
Blocking exists and works. Circumventing a block to keep contacting someone is itself a breach of this rule.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
- Not permitted: replying to every chapter of an author's novel with remarks about their appearance or intelligence.
- Permitted: a detailed negative review explaining why the prose does not work.
CG-CONDUCT-02 — Do not attack people on the basis of who they are.
Material dehumanising or inciting hatred against people for their race, ethnicity, national origin, religion, caste, disability, disease, sex, gender, gender identity, or sexual orientation is prohibited.
This is a rule about text addressed to readers. A character who holds vile views is a character; a comment expressing those views about real people is a breach.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-CONDUCT-03 — Do not publish private information about someone without their consent.
Home address, workplace, telephone number, government identifiers, financial details, medical information, and private images are all covered. So is aggregating public fragments into a profile that exposes someone.
Naming a pseudonymous author's legal identity against their wishes is covered by this rule.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-CONDUCT-04 — Do not organise or join a pile-on against a person or a work.
Recruiting others to mass-report, mass-downvote, or flood someone's comments is a breach for the organiser and for each participant.
Coordinated signal manipulation may also have its signals invalidated, which removes the effect without treating every participant as an equal offender.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, invalidate manipulated signals.
Authenticity and platform integrity
NovelVerse's discovery surfaces run on signals from real readers. These rules protect the meaning of those signals, and the ability to tell who you are talking to.
CG-AUTH-01 — Do not impersonate a person, an author, or NovelVerse itself.
Using someone's name, likeness, or author identity in a way designed to be mistaken for them is prohibited. Clearly-labelled parody and commentary are not impersonation.
Claiming to speak for NovelVerse or its operator is treated as impersonation.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-AUTH-02 — Do not create a new account to evade a restriction.
A suspension applies to the person, not to the account. Returning under a new identity while a restriction is in force extends it.
Possible outcomes after human review: restrict the account or feature, ban the account.
CG-AUTH-03 — One person, one account, where extra accounts would gain you something.
Additional accounts used to inflate a ranking signal, to obtain additional AI allowances, to vote on your own work, or to evade a limit are prohibited. This mirrors the Terms.
Possible outcomes after human review: restrict the account or feature, ban the account, invalidate manipulated signals.
CG-AUTH-04 — Do not manipulate rankings, reads, votes, or reviews.
Automated reads, purchased engagement, reciprocal vote rings, and review trading all misrepresent what readers actually did.
Where a campaign is found, the signals it produced are invalidated. That is a correction to the record rather than a punishment of everyone involved, and it can be applied without any account being restricted.
Possible outcomes after human review: invalidate manipulated signals, issue a warning, restrict the account or feature, ban the account.
CG-AUTH-05 — Do not post spam, scams, or phishing.
Bulk unsolicited promotion, links designed to harvest credentials or payments, malware, and schemes promising returns are removed on sight.
NovelVerse never asks for payment. Anything on this platform that appears to is a scam, and reporting it is the right response.
Possible outcomes after human review: remove the content, restrict the account or feature, ban the account.
CG-AUTH-06 — Do not scrape, bulk-extract, or work around platform limits and abuse controls.
Automated collection of other people's work, circumventing rate limits or AI allowances, and probing for unauthorised access are prohibited.
Security research is welcome, but it goes through the security contact rather than through production accounts and other people's data.
Possible outcomes after human review: restrict the account or feature, ban the account.
Rights, attribution, and AI assistance
You keep ownership of what you write here. These rules are about not taking what belongs to someone else, and about being straight with readers regarding how a chapter was made.
CG-RIGHTS-01 — Publish only what you have the right to publish.
This covers text, cover images, avatars, and audio. Work you did not create and are not licensed to use does not belong here, and neither does a translation published without the rights to make one.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-RIGHTS-02 — Passing off someone else's writing as your own is a breach of these Guidelines.
Plagiarism is handled here, as a community rule, and is not the same thing as a copyright claim. A rights holder who wants content disabled uses the Copyright Notice process; a reader who has spotted copied work uses the ordinary Report path.
Quotation, homage, shared tropes, and fan fiction are not plagiarism. Reproducing another author's expression and presenting it as your own is.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
CG-RIGHTS-03 — Declare AI assistance honestly when you publish a chapter revision.
When you publish or republish a chapter you state whether more-than-trivial model-written expressive text remains in that exact revision. If it does, the chapter carries an AI-assisted label. The declaration covers Nova and any external tool.
Proofreading, grammar correction, brainstorming, outlines, and suggestions need no label when no model-written prose remains in what you published.
NovelVerse runs no AI detector and publishes no prompts, providers, percentages, or private history. The declaration is yours, and a knowingly false one is reviewed by a person under this rule. It never triggers an automatic takedown.
Possible outcomes after human review: issue a warning, restrict the account or feature, remove the content.
CG-RIGHTS-04 — Do not republish material around a moderation removal or a copyright hold.
Re-uploading removed material, or publishing a lightly altered successor to it, extends the original decision rather than resetting it.
Where content is disabled by a copyright hold, the route back is the counter-notice process, not a second upload.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature, ban the account.
Mature content and labelling
Adult material has a place here. What these rules require is that readers can tell what they are walking into, and that the platform's own surfaces do not put it in front of someone who did not ask.
CG-LABEL-01 — Rate and tag your work accurately.
Set the maturity rating your content actually warrants, and keep it accurate as the work develops. Under-rating to reach a wider audience is a breach of this rule.
Where a rating is wrong, the ordinary correction is a rating change rather than removal.
Possible outcomes after human review: issue a warning, remove the content, restrict the account or feature.
CG-LABEL-02 — Covers, avatars, and banners are seen without consent, so keep them safe for every reader.
Images that appear on browsing surfaces are held to a stricter standard than the chapter behind them. Sexual imagery, graphic gore, and shock content are not permitted as covers or profile images regardless of the work's rating.
Uploaded images are screened automatically before publication. Automated screening is imperfect, which is what the appeal route is for.
Possible outcomes after human review: remove the content, issue a warning, restrict the account or feature.
CG-LABEL-03 — Do not spoil a work outside a marked spoiler.
Comments and reviews that reveal later events without marking them break the reading experience the platform exists for. Repeated deliberate spoiling is treated as harassment under CG-CONDUCT-01.
Possible outcomes after human review: remove the content, issue a warning.
What happens when a rule is broken
Decisions are made by a person. There is no automatic penalty in this system: automated screening can hold an image before publication and automated signals can raise something for review, but a restriction on an account or a removal of content is recorded by a human as a Verdict, with a notice explaining it.
The available outcomes are: a warning, which records the decision without limiting anything; removal of the specific content; restriction, which limits what an account can do while leaving reading, export, deletion, and appeal available; suspension, which ends access; and invalidation of manipulated signals, which corrects the record without penalising an account.
Reach is separate from removal. Material may be kept out of recommendation and discovery surfaces while remaining readable to anyone who goes to it directly. That is a narrower response than removal and it is used deliberately: the reach of a work and the right to publish it are different questions.
Severity, intent, history, and whether the breach targeted a person all bear on which outcome applies. The rules above list what each one can carry, not what it always does.
Reporting something
Report from the thing itself. Novels, chapters, and comments each carry a report control; profiles, chat messages, and direct messages carry their own. Choose the closest reason and add a short note — a note describing what to look at is worth more than a long one.
Reporting is not voting. Volume does not decide an outcome, and a report that turns out to be mistaken costs you nothing. Deliberately mass-reporting to suppress someone is a breach of CG-CONDUCT-04.
A rights holder asking for infringing material to be disabled needs the Copyright Notice process rather than a report; a report cannot carry the declarations that process requires.
Challenging a decision
Every adverse decision comes with a notice naming the rule identifier, the version of these Guidelines it was decided against, and a link to appeal. Appeals stay reachable while an account is restricted — that is deliberate, since an account that cannot appeal is one that cannot be wrong about.
A suspended account reaches its appeal through the link in its notice, which works without signing in.
An appeal is read by a person. It can restore removed content, lift a restriction, or leave the decision in place with an explanation.
How these Guidelines change
Each published version carries a version identifier and an effective date. A moderation notice cites the version in force when the decision was made, so a rule can be read as it stood at the time rather than as it stands now.
Rule identifiers are permanent. They are never renumbered and never reused; a rule that stops applying is marked retired and keeps its identifier so past notices remain readable.
A clarification — an example, a rewording, a better explanation of an existing line — advances only these Guidelines. A change to what is actually required of you also changes the Terms of Use, and you are asked to accept those again before continuing.