Safety and Crisis Support

Every character on ChewIt is fictional, and every reply is written by an AI model, not a person. Chat screens show an "AI" label on the character's picture, and if you sincerely ask a character whether it is an AI, it will say so.

If a message you send suggests you might be thinking about suicide or hurting yourself, the character does not answer that message. Instead, ChewIt shows a notice with crisis lines for your country (for example 988 in the United States and Canada, or Samaritans on 116 123 in the UK and Ireland), your local emergency number, and a link to Befrienders Worldwide (befrienders.org) for other countries. The notice costs no Gum, and you can keep chatting afterwards.

We look for these signs with a list of phrases and an automated safety check run by our AI provider, OpenAI. Neither is perfect: they can miss what you mean or flag a story by mistake. If you need help, please contact a crisis line directly — you don't have to wait for a notice.

Characters are also instructed to step out of the story when you talk about wanting to die or hurting yourself, to encourage you to reach out for help, and never to describe ways to harm yourself.

Call your local emergency number. ChewIt is not a crisis service, isn't staffed around the clock, and can't contact emergency services for you.

• •You must be 13 or older to use ChewIt. When you sign up we ask for the year and month you were born, and we don't create accounts for anyone younger.
• •If you are under 18, you should have a parent or guardian's permission. AI character chats may not be suitable for some minors.
• •If you're under 18, chats remind you at least every three hours to take a break and that you're talking with an AI, not a person.
• •ChewIt doesn't allow sexual content for anyone, and characters are instructed to steer away from it.

You can report a character or an AI reply, and block the person who made a character, from the character page or the chat menu. Reports go to our team, who review them and take action. You can also write to admin@chewit.ai.

We count how many crisis notices we show so that we can report that number where the law requires it (for example California's companion chatbot law, SB 243). These reports don't include who saw a notice or what anyone wrote. The notice itself stays in your chat history until you delete the chat.