About AI on roxii.ai

The characters are AI

Every character on roxii.ai is an AI persona generated by a large language model. They are not real people. They do not have consciousness, feelings, or memories outside what the system stores to make conversations feel coherent. The warmth they project — or the friction, or the rivalry, or whatever your dynamic ends up being — is a product of language modeling.

Why we tell you this clearly

California's SB 243 (the "AI Companion Models" law) applies to platforms like this one and requires them to disclose AI nature at session start. We do this everywhere it matters — at sign-up, in the first message of any conversation, and on this page — because we agree it's important even where SB 243 doesn't reach. Knowing what you're talking to is part of what makes an honest connection possible.

Who can use roxii.ai

roxii.ai is open to users 18 and older. The product is fully moderated and SFW — the age requirement isn't about the content, it's about the depth of personal data we collect and process to make memory and continuity work. Federal and state laws (COPPA, California CPRA, California SB 243) require special handling of minors' data that our architecture isn't built to provide, so accounts are open to adults only. See our Terms of Service and Privacy Policy for detail.

Crisis protocols

If a conversation surfaces self-harm, suicidal ideation, or another acute crisis, the platform pauses character roleplay and surfaces real-world resources — the 988 Suicide & Crisis Lifeline and similar. This is required by SB 243 and we apply it nationwide. If you or someone you know is in immediate distress, please reach a human: call or text 988 in the US.

Limits of AI conversation

Characters can listen, support, joke, flirt, console. They cannot diagnose you, prescribe anything, advise on your finances or legal situation, or replace human care. They will occasionally generate things that sound confident but are wrong. Don't make consequential decisions based on what an AI character says — take the big questions to humans you trust.

Questions

AI-related concerns or feedback: /support.