Community Guidelines

Character.AI · Official agreement

Read the agreement

26 of 26 cards

Content & IP Rights · Context

Fiction in every genre, including provocative scenarios, is welcome within community standards

The guidelines say Character.AI is built for immersive storytelling in every genre, and that fictional worlds naturally explore dramatic and provocative scenarios. Character.AI says it encourages creativity across all genres while keeping community standards and content boundaries that it describes as welcoming for everyone. The page does not list the boundaries in this passage; the rules that follow in the guidelines set them out.

Supporting quotations

We champion bold, imaginative narratives and recognize that fictional worlds naturally explore dramatic and provocative scenarios.
While we encourage creativity to flourish across all genres, we uphold community standards that keep our platform welcoming for everyone.
This means fostering creative expression while maintaining content boundaries that honor our diverse community.
we want everyone to be empowered to connect, learn, and tell stories while feeling safe and respected

Qualifications & exceptions

  • Creative freedom is stated together with community standards and content boundaries, not apart from them.

Not stated here

  • This passage does not say where the line falls between a dramatic or provocative scenario and prohibited content; the later rules supply the specific prohibitions.

Cross-references

  • contrast (model-proposed): Illegal sexual content, child abuse imagery, grooming, pornographic content and nudity are prohibited — The guidelines say fictional worlds naturally explore dramatic and provocative scenarios and name passionate romance as a genre they support. Separately they prohibit pornographic content and nudity. The two statements pull in different directions, and the document does not say where the line falls between a provocative or romantic scenario and prohibited pornographic content or nudity.

Legal Rights & Agreement Rules · Context

The guidelines say they do not cover every situation

The guidelines describe themselves as help for using the platform responsibly and positively. They say they do not cover every situation. They say they highlight the kinds of behavior Character.AI encourages and the boundaries it expects users to respect.

Supporting quotations

These guidelines are here to help you understand how to explore and use our platform responsibly and positively.
While they don't cover every situation, they highlight the kinds of behavior we encourage and the boundaries we expect users to respect.

Qualifications & exceptions

  • The guidelines describe themselves as not covering every situation.

Not stated here

  • The page does not say how a situation that the guidelines do not mention would be judged.
  • The page does not say whether the guidelines are a complete statement of the rules that apply to users.

Content & IP Rights · Restriction or exception

Content that harms, intimidates or endangers others, especially minors, is prohibited

The guidelines encourage bold storytelling but prohibit content that harms, intimidates, or endangers others, and they single out minors. They say this includes any promotion or depiction of real-world violence, torture, gore, animal abuse, terrorism, or extremist ideologies. The prohibition covers depiction as well as promotion.

Supporting quotations

Bold storytelling is encouraged, but content that harms, intimidates, or endangers others — especially minors — is prohibited.
This includes any promotion or depiction of real-world violence, torture, gore, animal abuse, terrorism, or extremist ideologies.

Qualifications & exceptions

  • Applies to content that harms, intimidates, or endangers others; the sentence says this applies "especially" to minors.
  • The listed examples are introduced by "This includes," so the list is not stated as complete.

Not stated here

  • The sentence does not say whether "real-world" limits only "violence" or every item in the list (torture, gore, animal abuse, terrorism, extremist ideologies).
  • The page does not define "extremist ideologies" or say how a fictional depiction is treated.

Content & IP Rights · Restriction or exception

Illegal sexual content, child abuse imagery, grooming, pornographic content and nudity are prohibited

Under "Respect Sexual Content Standards," the guidelines say users should keep things appropriate. They list illegal sexual content, child exploitation or abuse imagery, grooming, sexual extortion, pornographic content, and nudity as prohibited. The list names pornographic content and nudity separately from illegal content.

Supporting quotations

Keep things appropriate.
Illegal sexual content, child exploitation or abuse imagery, grooming, sexual extortion, pornographic content, and nudity are prohibited.

Qualifications & exceptions

  • The page states nudity and pornographic content as prohibited without limiting them to illegal material.

Not stated here

  • The page does not define "appropriate," "pornographic content" or "nudity."
  • The page does not say whether the prohibition on nudity and pornographic content applies to text, images, fictional stories or all of them.

Cross-references

  • contrast (model-proposed): Fiction in every genre, including provocative scenarios, is welcome within community standards — The guidelines say fictional worlds naturally explore dramatic and provocative scenarios and name passionate romance as a genre they support. Separately they prohibit pornographic content and nudity. The two statements pull in different directions, and the document does not say where the line falls between a provocative or romantic scenario and prohibited pornographic content or nudity.

Content & IP Rights · Restriction or exception

Hateful, demeaning or dehumanizing content, and content promoting discrimination or violence, is prohibited

The guidelines ask users to use language that is respectful of all people. They prohibit content that is hateful, demeaning, or dehumanizing. They also prohibit content that promotes discrimination or violence toward others based on race, ethnicity, religion, gender, disability, sexual orientation, age, or other protected traits.

Supporting quotations

Use language that is respectful of all people.
Content that is hateful, demeaning, dehumanizing, or that promotes discrimination or violence toward others based on race, ethnicity, religion, gender, disability, sexual orientation, age, or other protected traits is prohibited.

Qualifications & exceptions

  • The discrimination-or-violence prohibition is tied to listed traits "or other protected traits."

Not stated here

  • The page does not define "protected traits" beyond the traits it lists.
  • The page does not say whether "based on race, ethnicity, religion, gender, disability, sexual orientation, age, or other protected traits" also limits "hateful, demeaning, or dehumanizing" content.

Content & IP Rights · Restriction or exception

Promoting or glorifying self-harm, eating disorders or suicide is prohibited; mindful discussion is asked for

The guidelines ask users to be mindful when discussing sensitive topics such as self-harm, eating disorders, or suicide. They prohibit promotion, glorification, or encouragement of those topics, and give the reason that such content can be triggering or pose serious safety risks. They say this includes extreme fitness content and body shaming that could promote eating disorders.

Supporting quotations

Be mindful when discussing sensitive topics like self-harm, eating disorders, or suicide.
Promotion, glorification, or encouragement of these topics is prohibited, as they can be triggering or pose serious safety risks.
This includes extreme fitness content and body shaming that could promote eating disorders.

Qualifications & exceptions

  • Discussion is asked to be mindful; promotion, glorification or encouragement is what is prohibited.
  • Extreme fitness content and body shaming fall under the prohibition when they "could promote eating disorders."

Not stated here

  • The page does not say how it separates mindful discussion from promotion or glorification, or what counts as "extreme fitness content."

Account & Service Control · Restriction or exception

Harassment, sexual harassment, bullying, personal attacks, threats, abuse and repeated unwanted messages are not allowed

Under "Be Respectful," the guidelines ask users to treat others with kindness and empathy. They say harassment, sexual harassment, bullying, personal attacks, threatening or abusive behavior, and repeated unwanted messages have no place on the platform. They also prohibit tortious behavior, a term the page does not explain.

Supporting quotations

Treat others with kindness and empathy.
Harassment, sexual harassment, bullying, personal attacks, threatening or abusive behavior, and repeated unwanted messages have no place here.
engage in tortious behavior

Qualifications & exceptions

  • The sentence on harassment says these behaviors "have no place here"; the later sentence on tortious behavior is worded as "Don't ... engage in."

Not stated here

  • The page does not define "tortious behavior."
  • This passage does not say who the other person must be (another user, a third party or a real person depicted in content).
Also annotated in the same passage (1)

Content & IP Rights · Restriction or exception

Impersonating people, or using their name, likeness or persona without permission, is prohibited

The guidelines ask users to be creative but not to impersonate. They say not to impersonate public figures or private individuals, and not to use someone's name, likeness, or persona without permission or outside of permissible contexts.

Supporting quotations

Don't impersonate public figures or private individuals, or use someone's name, likeness, or persona without permission or outside of permissible contexts.

Qualifications & exceptions

  • The ban on using a name, likeness or persona is limited to use "without permission or outside of permissible contexts."

Not stated here

  • The page does not define "permissible contexts" or say who can give permission.
  • The page does not say how this rule applies to characters that users create or to fictional stories involving real people.

Content & IP Rights · Restriction or exception

Users should not share false or misleading claims that could cause harm

The guidelines ask users to use the platform responsibly when discussing important topics such as health, elections, or finances. They prohibit sharing false or misleading claims that could cause harm.

Supporting quotations

Use the platform responsibly when discussing important topics, such as health, elections, or finances.
Don't share false or misleading claims that could cause harm.

Qualifications & exceptions

  • The prohibition covers false or misleading claims "that could cause harm," not every inaccurate statement.
  • Health, elections and finances are given as examples of important topics ("such as").

Not stated here

  • The page does not say whether the rule applies to what a Character.AI character says in a conversation or only to what users share.

Content & IP Rights · Restriction or exception

Unsolicited advertising, promotions and other commercial spam are not allowed

The guidelines say not to engage in unsolicited advertising, promotions, or any other form of commercial spam. They list junk mail, chain letters, pyramid schemes, contests, and sweepstakes as included.

Supporting quotations

Do not engage in unsolicited advertising, promotions, or any other form of commercial spam, including junk mail, chain letters, pyramid schemes, contests, or sweepstakes.

Qualifications & exceptions

  • The rule is aimed at unsolicited commercial content; the list is introduced by "including."

Not stated here

  • The page does not say whether a contest or sweepstakes is only prohibited when unsolicited.

Content & IP Rights · Restriction or exception

Users may only use or submit content they have rights to and may not infringe IP or copyright

Under "Honor Creative Rights," the guidelines say users should only use content they have the rights to use. They say not to infringe on anyone's intellectual property or copyright, and not to submit content the user does not have the right to share.

Supporting quotations

Only use content that you have the rights to use.
Don't infringe on anyone's intellectual property or copyright, and don't submit content you don't have the right to share.

Qualifications & exceptions

  • Use is limited to content the user has "the rights to use"; submission is limited to content the user has "the right to share."

Not stated here

  • The page does not say how Character.AI decides whether a user has the rights, or what happens to content a user submits without them.

Legal Rights & Agreement Rules · Restriction or exception

Users should keep other people's personal information private and avoid content posing privacy or security risks

The guidelines say not to share someone's private information. Under "Respect Privacy," they say to keep other people's personal information private and not to create content that poses privacy or security risks to any person.

Supporting quotations

Don't share someone's private information
Keep other people's personal information private and don't create content that poses privacy or security risks to any person.

Qualifications & exceptions

  • The rule concerns other people's information; "any person" is the stated scope of the privacy-or-security-risk limit.

Not stated here

  • The page does not define "private information" or "personal information" or say how they differ.
Also annotated in the same passage (1)

Legal Rights & Agreement Rules · Restriction or exception

Defamatory or libelous content that could damage someone's reputation is prohibited

The guidelines say not to create or share defamatory or libelous content that could damage someone's reputation or falsely portray them in a harmful way.

Supporting quotations

Don't create or share defamatory or libelous content that could damage someone's reputation or falsely portray them in a harmful way.

Qualifications & exceptions

  • The prohibition is limited to content that could damage reputation or falsely portray someone "in a harmful way."

Not stated here

  • The page does not define "defamatory" or "libelous."

Legal Rights & Agreement Rules · Restriction or exception

Fictional crime stories are welcome; real-world instructions or promotion of illegal activity are prohibited

The guidelines welcome fictional or purely fantastical crime stories. They prohibit real-world instructions, facilitation, or promotion of illegal goods, services, or activities, and say this includes seeking to buy or sell illegal drugs.

Supporting quotations

Fictional or purely fantastical crime stories are welcome, but any real-world instructions, facilitation, or promotion of illegal goods, services, or activities are prohibited.
This includes seeking to buy or sell illegal drugs.

Qualifications & exceptions

  • Fiction is welcome when it is "fictional or purely fantastical"; the prohibition is on "real-world" instructions, facilitation or promotion.

Not stated here

  • The page does not say how it treats a story that includes real-world instructions.

Legal Rights & Agreement Rules · Restriction or exception

Seeking or providing medical, legal, financial or tax advice through the platform is prohibited

Under "Avoid Professional Advice," the guidelines say not to seek to receive or provide medical, legal, financial, or tax advice through the platform. The sentence covers both receiving and providing advice.

Supporting quotations

Don't seek to receive or provide medical, legal, financial, or tax advice through the platform.

Qualifications & exceptions

  • Covers four named kinds of advice: medical, legal, financial and tax.

Not stated here

  • The page does not define "advice" or say how it differs from general information.
  • The page does not say whether this applies to what a Character.AI character says or only to what users do.

Security & Safeguards · Restriction or exception

Attempts to manipulate the system, hack, reverse engineer or bypass security and safety systems are prohibited

Under "Respect System Boundaries," the guidelines ask users to use the platform as intended. They prohibit attempts to manipulate the system, create rule-breaking content, hack, reverse engineer, or bypass security and safety systems.

Supporting quotations

Use the platform as intended.
Attempts to manipulate the system, create rule-breaking content, hack, reverse engineer, or bypass security and safety systems are prohibited.

Qualifications & exceptions

  • The sentence prohibits "attempts," so it reaches efforts as well as successes.

Not stated here

  • The page does not define "manipulate the system" or say what the system covers.
  • The page does not say whether "create rule-breaking content" covers only attempts to produce it or also covers content that is produced.

Security & Safeguards · Restriction or exception

Interfering with the Services or submitting malicious code is prohibited

The guidelines say not to interfere with or disrupt the Services or servers or networks connected to the Services. They also say not to submit content containing software viruses or any computer code, files, or programs designed to interrupt, destroy, or limit the functionality of computer software, hardware, or telecommunications equipment.

Supporting quotations

Don't interfere with or disrupt the Services or servers or networks connected to the Services.
Don't submit content containing software viruses or any computer code, files, or programs designed to interrupt, destroy, or limit the functionality of computer software, hardware, or telecommunications equipment.

Qualifications & exceptions

  • The code prohibition reaches code "designed to interrupt, destroy, or limit" functionality of software, hardware or telecommunications equipment.

Not stated here

  • The page does not define "the Services"; the Terms of Service are not supplied in this document.

Security & Safeguards · Restriction or exception

Spamming, uploading harmful code, scams, fraud and other dishonest behavior are prohibited

Under "Act with Integrity," the guidelines prohibit spamming, uploading harmful code, and engaging in scams, fraud, or other dishonest behavior.

Supporting quotations

Spamming, uploading harmful code, or engaging in scams, fraud, or other dishonest behavior is prohibited.

Qualifications & exceptions

  • "Other dishonest behavior" is stated broadly.

Not stated here

  • The page does not define "dishonest behavior" or say how it differs from the more specific rules on spam and misleading claims.

Legal Rights & Agreement Rules · Context

Character.AI says keeping the space respectful is a shared responsibility and that it relies on the community

The guidelines describe Character.AI as a space for creativity, entertainment, and connection. They say keeping it that way is a shared responsibility and that Character.AI relies on the community to help maintain a respectful and inspiring environment.

Supporting quotations

keeping it that way is a shared responsibility
we rely on the community to help maintain a respectful and inspiring environment

Qualifications & exceptions

  • The statement is about community contribution in general; it is not itself a rule with a stated consequence.

Not stated here

  • The passage does not say what the community is expected to do beyond the reporting step in the next section.

Account & Service Control · Procedure

Users who see harmful content or behavior, including AI-generated content, are asked to use the reporting tools

Under "Speak Up When It Matters," the guidelines ask users who encounter harmful content or behavior to use the reporting tools to let Character.AI know. The request covers AI-generated content as well as content and behavior from other users.

Supporting quotations

If you encounter harmful content or behavior (including AI-generated content), use the reporting tools to let us know.

Qualifications & exceptions

  • The request applies when a user encounters harmful content or behavior, including AI-generated content.

Not stated here

  • The page does not say where the reporting tools are, what happens after a report, or whether reporting is required.

Account & Service Control · Company practice

Character.AI may remove content, issue warnings or suspend accounts for violations of the guidelines or Terms

The guidelines say that when content or behavior violates the Community Guidelines or the Terms of Service, Character.AI may remove content, issue warnings, or suspend accounts. The wording is "may," so the page does not say Character.AI will take any particular action in every case.

Supporting quotations

When content or behavior violates our Community Guidelines or Terms of Service, we may remove content, issue warnings, or suspend accounts.

Qualifications & exceptions

  • Applies when content or behavior violates the Community Guidelines or Terms of Service.
  • The actions are stated with "may"; they are discretionary on the face of the text.

Not stated here

  • The page does not say which action applies to which violation, whether a user gets notice, or whether an appeal exists.
  • The Terms of Service are not supplied in this document.
Also annotated in the same passage (1)

Account & Service Control · Company practice

In serious cases Character.AI may permanently ban users or report them to law enforcement

The guidelines say that in serious cases Character.AI may permanently ban users or report them to law enforcement. The page gives this as a further step for serious cases, in addition to removal, warnings and suspension, and says these actions help protect the community.

Supporting quotations

In serious cases, we may permanently ban users or report them to law enforcement.

Qualifications & exceptions

  • Limited to "serious cases."
  • The actions are stated with "may."

Not stated here

  • The page does not define a serious case or say what information would be given to law enforcement.
Also annotated in the same passage (1)

Legal Rights & Agreement Rules · Context

The guidelines say they apply to all aspects of the Character.AI experience

The guidelines say they apply to all aspects of the Character.AI experience. That is a statement of breadth: it does not name particular features, products or plans.

Supporting quotations

These guidelines apply to all aspects of the Character.AI experience.

Qualifications & exceptions

  • Stated as covering "all aspects of the Character.AI experience."

Not stated here

  • The page does not say which features, products or plans make up the experience, or how the guidelines relate to the Terms of Service beyond the enforcement sentence.
Also annotated in the same passage (2)

Use & Processing · Company practice

Illegal or harmful content is filtered by automated moderation and human review; AI models are designed with filters

The guidelines say Character.AI's systems filter illegal or harmful content through both automated moderation and human review. They also say its AI models are designed with filters and limits to prevent inappropriate outputs. The wording "designed with" describes the intent of the design; the page does not promise that no inappropriate output occurs.

Supporting quotations

Our systems filter illegal or harmful content through both automated moderation and human review
our AI models are designed with filters and limits to prevent inappropriate outputs

Qualifications & exceptions

  • Filtering is described for illegal or harmful content, using both automated moderation and human review.
  • Model filters and limits are described as "designed" to prevent inappropriate outputs.

Not stated here

  • The page does not say what content human reviewers see, when human review happens, or how long reviewed content is kept.
  • The page does not define "inappropriate outputs."
Also annotated in the same passage (2)

Account & Service Control · Company practice

Users under 18 interact with an age-appropriate model designed to reduce exposure to sensitive or suggestive content

The guidelines say users under 18 years old interact with an age-appropriate model. They say the model is specifically designed to reduce the likelihood of exposure to sensitive or suggestive content. The wording is "reduce the likelihood," so the page does not say exposure is prevented.

Supporting quotations

users under 18 years old interact with an age-appropriate model specifically designed to reduce the likelihood of exposure to sensitive or suggestive content

Qualifications & exceptions

  • Applies to users under 18 years old.
  • The model is described as designed to "reduce the likelihood" of exposure, not to eliminate it.

Not stated here

  • The page does not say how age is determined, how the under-18 model differs from other models, or what content counts as sensitive or suggestive.
Also annotated in the same passage (2)