An assistant that runs on our servers, not on your phone.

That sounds like an implementation detail. It is the difference between a safe design and a common, bad one.

What it does

A way into the product, not a replacement for it.

Takes a question asked in ordinary language and searches CHiiRO for it.
Points to where in CHiiRO a particular kind of conversation is happening — communities, discussions, questions and people.

The most valuable thing on CHiiRO is still another person who has already done the thing you are about to do. The assistant’s job is to shorten the path to that person, not to stand in for them.

What it deliberately is not

On the near side of a real line, on purpose.

It is not a persona, and it does not pretend to be a person.
It is not a mentor, an advisor or a coach, and it will not be marketed as one.
It is not a conversational character.

Conversational characters are regulated territory, and the questions there deserve a proper answer rather than a fast one.

Current state

Where it runs: On CHiiRO’s servers, which send your question to an AI provider; the AI credential is not in the app
Where you reach it: The Android app on Google Play
What you type: Handled as the Privacy Policy describes
Conversational characters: Not enabled

“Your key never leaves our servers” is a security claim, so it is stated only as far as it is true: the credential is on our side. What you type is not kept out of the logs, and this readout says so.

Read the disclosures

The mistake almost everyone makes.

The quickest way to put an AI assistant in a mobile app is to embed the credential in the app and let the phone call the AI provider directly. It works on the first try, it ships in an afternoon, and it means the key is now sitting inside an artefact that anyone can download and take apart. Keys extracted this way get abused at someone else’s expense, usually for weeks before anyone notices.

The design CHiiRO built does not do this: your phone talks to CHiiRO, and CHiiRO talks to the assistant, so the credential stays on our side.

How your data is protected

What happens to what you type.

Your question goes to CHiiRO’s servers, which ask an AI provider for the answer. It is handled as the Privacy Policy describes.

How search reads a plain question

Questions about the assistant.

How AI-written content is recorded

Does CHiiRO have an AI assistant?

Yes. The Android app on Google Play has an assistant you can ask in ordinary language; it searches CHiiRO for you. It runs on CHiiRO’s servers, so the AI credential stays off your device; to answer, those servers send your question to an AI provider (OpenAI or Microsoft Azure OpenAI Service).

What happens to what I type into the CHiiRO assistant?

It is sent to CHiiRO’s servers, which ask an AI provider for the answer, and it is handled as the CHiiRO Privacy Policy describes.

Can I talk to an AI mentor on CHiiRO?

No. The assistant helps you use the network; it is not a persona, a mentor or a coach and does not present itself as a person. CHiiRO has not enabled conversational AI characters, because the age-rating and regulatory questions there need answering first.

Why does it matter where an AI assistant runs?

Because an app that calls an AI provider directly must carry the credential inside the app, where anyone who downloads it can extract it. Brokering server-side keeps the credential off the device. It is a common shortcut and a consequential one.