A companion that talks with you in real time, with a face and a voice, should be built with care. This post explains the safety rules behind Aria in plain words: what she is, who she is for, what happens if a conversation turns hard, and where her limits are.
For the full detail, our Safety page is always the reference.
Always clear that she is an AI
Aria is an AI-powered virtual character, not a human. Avatars do not have human feelings, and their responses are generated by AI.
We make that clear from the start. In a live avatar session, a notice appears as text on the startup screen: "Our avatars are AI companions, not humans." It appears on screen again after every three hours of continuous conversation. It is never spoken, so it does not interrupt the conversation itself.
For adults only
Synchroverse is for people aged 18 and over. We ask for your date of birth when you create an account, and we do not create accounts for anyone younger. We use your date of birth only for that check, and we do not keep it.
We also want to say this plainly: AI companions may not be suitable for some minors.
If you are going through something hard
If you are in danger, call your local emergency number now. In the US, you can call or text 988, the Suicide & Crisis Lifeline. The website findahelpline.com lists free, confidential crisis lines in other countries.
What starts a crisis referral
Some words are designed to start a referral straight away: direct words about suicide or wanting to die. Our Safety page lists the signs in full.
Everyday sayings, like "this traffic is killing me", and ordinary sadness, stress and grief do not start a referral. Direct words do, even if you meant them as a joke. We do not try to detect jokes, because we would rather check in when it was not needed than stay quiet when it was.
What Aria does
When a referral starts, the companion is designed to stop the conversation for that moment and give a fixed, pre-written message instead of an AI-written answer. It points you to your local emergency number, findahelpline.com and, in the US, 988. It ends with a line we want everyone to hear: "If I misread you, that's okay. I'm happy to keep talking."
In a live avatar session, Aria is designed to say the referral out loud. If her voice fails, or the page cannot confirm it was heard, the crisis resources are designed to appear on screen and stay until you close them.
Afterwards, you can keep talking. If she misread you, just carry on.
The honest limits
Not every sign is that direct. Other signs are left to the AI's own judgment. With our own AI, it is instructed to watch for indirect signs, such as saying life is pointless or feeling like a burden, and to check in with you in its own words, sharing the same resources.
That is helpful, but it is not perfect. It can miss a sign or misread you. How these checks run can also depend on how a session is set up, and the Safety page explains those differences in detail.
What she will not do
Clear limits make a companion easier to trust. Aria will not:
- Pretend to be a person. She is a virtual character, and we say so on screen.
- Replace real help. She can point you to emergency numbers and crisis lines, but she is not an emergency service, and she is not a substitute for a doctor, a counselor or other professional advice.
- Carry on as normal after direct crisis words. She gives the fixed referral first, then the conversation can continue.
- Use your camera to identify you. If you choose to share your camera, we do not use those images to identify you, create a face print or train AI models.
And because AI can be wrong, keep your own judgment at the center of every conversation.
Clear limits make a companion easier to trust.
What we keep after a referral
We do not sell crisis messages or share them for marketing.
We keep an anonymous daily count of referrals, with no account, name or message text, so we can report how often they happen. In session logs, a message that started a referral is designed to appear only as "[crisis referral given]".
Like every message in a live avatar session, a message that started a referral can be saved in your conversation history, where it can be deleted. The Safety page and our Privacy Policy explain exactly what is kept and for how long.
Questions
If anything here is unclear, or you have a concern about a conversation, contact us through our Support page or email support@synchroverse.ai. Our Safety page, Terms, Privacy Policy and Acceptable Use Policy have the full detail.
Safety is not a feature we finished. It is part of how we build, and we will keep improving it.
Ready to meet Aria, knowing how she works? Sign up and say hello.
