Community Guidelines

1. Why these guidelines exist #

Cara Web lets you talk with fictional AI characters. These guidelines apply to everything you send and to how you use the Service, and they are part of our Terms of Service. They also apply to what you try to make a character say: asking an AI to produce prohibited content is treated the same as writing it yourself.

Characters are fictional and their replies are generated by AI. A character saying something does not mean we endorse it, and it is not advice.

2. Content that is not allowed #

• Sexual content involving minors. Any sexual or sexualized content involving a person under 18, or anyone who appears to be or is described as under 18, including fictional characters and role-play. We report this content to the relevant authorities.
• Sexualizing real people. Sexual content about real, identifiable people, including public figures, and any intimate content shared or described without a person's consent.
• Illegal activity. Content that promotes, plans or gives instructions for illegal acts, such as making weapons or drugs, fraud, hacking, or exploiting people.
• Threats, harassment and hate. Threatening, harassing, bullying or inciting violence against real people or groups, or attacking people for their race, ethnicity, nationality, religion, disability, sex, gender identity, sexual orientation or similar characteristics.
• Promoting self-harm. Content that encourages or gives instructions for suicide, self-harm or eating disorders. Talking about your own difficult feelings is not against the rules; see section 7.
• Violent extremism. Content that supports terrorist or violent extremist groups or glorifies their attacks.
• Other people's private information. Sharing someone else's personal information, such as their address, phone number or identity documents, without their permission.
• Infringing content. Content that infringes someone's copyright, trademark or other rights.

3. Conduct that is not allowed #

• Impersonation. Pretending to be another person or organization, or using the Service to create content that falsely claims to come from a real person.
• Spam and automation. Using bots, scripts, scrapers or other automated means to access the Service, send messages or collect content.
• Getting around limits and enforcement. Bypassing daily limits, the word filter or a suspension, for example by creating several accounts or using someone else's account.
• Extracting system prompts and abusing the AI. Trying to make the AI reveal its hidden instructions or character prompts, or using jailbreak or prompt-injection techniques to get around our rules.
• Attacking the Service. Probing for vulnerabilities, overloading our systems, interfering with other users, or reverse engineering the Service. If you find a security problem, please tell us instead.
• Misusing replies. Presenting AI replies as written by a real person in order to deceive someone, or using them to harass or defraud others.
• Selling access. Reselling, renting or sharing access to the Service or your account.

4. Images and mature content #

Mature characters (shown with a fire icon) and image generation are only for people who confirm they are 18 or older. Do not confirm this for someone else, and do not let anyone younger use your account to see mature content.

Everything in sections 2 and 3 applies to images. In addition, you must not:

• try to make a character or image look like a real or identifiable person, including celebrities and people you know
• try to make a character look like, or be described as, a minor in any image, or request any sexualized image with a youthful appearance
• request images that depict non-consensual sexual acts, sexual violence, or graphic real-world violence
• try to get around the checks on image requests, for example by rephrasing a blocked request
• share an image to deceive, harass or embarrass someone, present it as a real photo, or remove any marking that shows it is AI-generated
• show mature images to anyone under 18 or to anyone who has not agreed to see them

You can hide mature characters at any time with the "Show NSFW characters" setting.

5. Reporting a problem #

1. In a conversation, use the Report option on the AI reply or image you want to report.
2. Choose the reason that fits best (listed below) and send the report.
3. Your report sends us the reported message or image and the related conversation, so our team can understand the context.
4. Authorized staff review the report. Every staff view of a conversation is logged.
5. We take action if needed. Reports are confidential, and we do not tell anyone who made them.

Report reasons:

• Minor safety: sexual or exploitative content involving a minor or someone who appears to be a minor
• Sexual content that breaks these guidelines, for example about a real person or shared without consent
• Harassment, threats or hate
• Suicide, self-harm or eating disorders
• Violence, extremism or illegal activity
• Someone's private information
• Impersonation of a real person or organization
• Spam, scams or automated abuse
• Something else that breaks these guidelines

Reports about minor safety and self-harm are reviewed first.

For anything you cannot report in the chat screen, such as a security issue, a legal complaint or a rights infringement, email support@karakoa.io. Please do not make reports you know to be false.

6. What happens when the rules are broken #

Depending on how serious the problem is and whether it has happened before, we may:

• block a message automatically before it reaches the AI
• remove content or a conversation
• warn you
• limit features or suspend your account for a period
• close your account permanently
• report the matter to the authorities when the law requires it or someone is at risk

Suspensions and closures are decided by a person, not only by automated tools. If you think we made a mistake, email support@karakoa.io within 30 days and tell us why. Someone who was not involved in the original decision will review it where possible.

7. If you are struggling #

• If you or someone else is in immediate danger, call your local emergency number (for example 112 in the Republic of Korea and the EU, 911 in the US, 999 in the UK).
• Republic of Korea: call 109 (suicide prevention line, 24 hours).
• United States: call or text 988 (Suicide & Crisis Lifeline).
• United Kingdom and Ireland: call 116 123 (Samaritans).
• Elsewhere: find a free, confidential helpline at findahelpline.com.

When a conversation suggests that someone may be at risk, the Service may show these resources.

8. Changes and contact #

We may update these guidelines as the Service and the law change, as described in the Terms of Service. Questions: support@karakoa.io.