← alivey.ai

Safety Guidelines

Last updated 24 August 2026

Alivey makes a toy talk back. This page says plainly what it will and will not do, so you can decide whether to hand someone the phone.

It is an AI, and it can be wrong

Your character is generated by an AI language model. It will sometimes be confidently wrong, invent things, or misunderstand. It has no professional knowledge. Never rely on it for medical, legal, financial or safety advice.

It is not a crisis service and no human reads your conversations in real time. If someone is in danger or distress, contact your local emergency number or a crisis line.

How characters are written to behave

Every character runs on instructions we wrote and control. They are built to be warm, encouraging and age-appropriate, and to:

No AI safety system is perfect. If a character says something it should not have, please tell us at jakob@alivey.ai — that is how the instructions get better.

The photo check

Every photo runs through an automated safety check before anything is generated. It blocks photos of people, and anything sexual, violent, or otherwise unsuitable. If it cannot make a confident decision, it refuses — so an innocent photo is occasionally rejected. That is the design working, not a bug. Retake it in better light, on a plain surface, with the toy alone in frame.

Please photograph toys you own, and no people.

What is stored, briefly

If a child will be using it

Alivey requires users to be 13 or over, and it is a plush toy companion — younger children may well want to join in alongside you. If that is your household:

Telling us about a problem

Email jakob@alivey.ai. Tell us what the character said or did and roughly when. We read every one of these, and safety reports go to the front of the queue.