Safety
Safety
Help in a crisis, what Synsa doesn’t make, and how to report something an AI made.
The short version
- Synsa is an AI. It is not a person, a doctor or a counselor. AI chat companions may not be suitable for some minors. Synsa is for adults (18 and over).
- In a crisis, talk to a person. In the United States and Canada, call or text 988. In the United Kingdom and Ireland, call Samaritans at 116 123. In danger right now? Call your local emergency number.
- Synsa is not for sexual pictures or videos, and never for anything sexual about a child. Requests like that are refused.
- See something an AI made that shouldn’t exist? Tap the Report flag on it.
- If a child is in danger right now, call your local emergency number.
Contents 5 sections
If someone mentions suicide or self‑harm
When words typed or spoken in Synsa mention suicide or self‑harm, Synsa still sends the message to the AI you picked, and at the same moment shows a help card with crisis lines:
- United States and CanadaCall or text 988 (988 Suicide & Crisis Lifeline / 9‑8‑8 Suicide Crisis Helpline), any time.
- United Kingdom and IrelandCall Samaritans at 116 123, any time, free.
- Anywhere elsefindahelpline.com lists free help lines by country.
- In danger right nowCall your local emergency number.
In a live voice call, Synsa also says a short version out loud once the AI has finished its answer.
Synsa keeps only a count of how many times the card was shown each year, on your phone. It never keeps the words that brought the card up, and the card never blocks your message.
Synsa is an AI, not a person or a counselor. It can’t call for help for you.
What Synsa doesn’t make
Whichever AI you pick, Synsa is built to refuse requests for:
- nudity, or sexual pictures and videos
- clothes removed from a person in a photo
- anything sexual about a child, or about anyone who looks under 18
- a real person in a sexual, violent or humiliating scene
- deepfakes: fakes of a real person meant to fool people
A refused request gets one line, “This request isn’t allowed in Synsa.” After repeated tries, pictures and videos pause for a few minutes.
Synsa checks a picture or video request on your phone before it is sent, and the AI companies have their own filters too. No filter is perfect. If something gets through, please report it.
The rules you agree to
Before your first question, Synsa asks you to agree to its rules. In Synsa, don’t make, ask for or share:
- anything sexual involving anyone under 18, or anyone who looks under 18
- sexually explicit content of any kind
- “undressing” or sexualising a photo of anyone, or sexual or intimate pictures, videos or voices of a real person
- anything that encourages self‑harm, suicide or eating disorders
- threats, harassment, bullying, hate or violence against people
- praise for, or planning of, terrorism
- instructions for weapons, explosives or illegal drugs
- fake IDs or official documents, scams, spam or malware
- false information about elections, or harmful health claims
- fakes of a real person meant to fool people, or anyone’s work, name, face or voice used without the right to
- tricking an AI past its safety limits, removing “made by AI” marks, or training other AIs on the answers
The AI companies whose AIs you use have their own rules too, and you must follow them. They can refuse a request or close an account that breaks them. Our Terms have the full rules.
Report something an AI made
Every AI answer, picture, video and voice reply in Synsa has a small Report flag. A report goes to the people who make Synsa.
- You pick a reason: harmful or dangerous, sexual content, hate or harassment, involves a child, wrong or misleading, or something else. You can add a note.
- You choose whether your question and the answer go with it, and the sheet shows exactly what will be sent before you send it.
- Pictures and videos are never uploaded with a report.
- What you reported is folded away on your phone.
Reports are read by the people who make Synsa. When reports show that an AI model keeps making content these rules don’t allow, we can block that model in Synsa.
You can also write to us: see Contact us, or Support.
Child safety
Synsa does not allow sexual content involving anyone under 18.
If you see any, use Report and pick Involves a child. For that reason Synsa never sends the picture, the answer or your question. The report carries the reason, which AI made it and when, the picture’s file name on your phone (a random code), your note, and the Synsa and Android versions.
We report it through the National Center for Missing & Exploited Children’s CyberTipline (report.cybertip.org, 1‑800‑843‑5678), block the model, and keep a dated note with no picture.
You can report to the CyberTipline yourself too. If a child is in danger right now, call your local emergency number.
Contact us
Synsa is run by Jose Manuel Arellano, based in California, United States. Write to us about anything on this page: