AI Transparency
Businesses use Neaion to answer the messages their customers send them. If you have written to one of those businesses, some or all of the reply may have been written by AI. This page explains what that AI does, what it will not do, and how a person stays in charge.
It is an AI, and it says so
Neaion writes in a natural, human tone, but it never claims to be a person. If you ask whether you are talking to an AI, it answers yes, plainly.
It answers from the business's own rules — and does not invent
Replies are grounded in what the business has told Neaion: its policies, prices, opening hours and saved answers. Before a reply is sent, a separate check confirms the business's rules actually cover the question. When they do not, Neaion does not guess. It tells you it will check and come back — and it asks a person at the business.
A person decides what it cannot
When Neaion asks the business a question, the owner's answer is what comes back to you. It is rewritten in your language and checked, before it is sent, to make sure nothing was added that the owner did not say.
It acts within limits it cannot change
Everything Neaion proposes to do passes a rule engine the business configures and the AI cannot override. It can only reply to the person who wrote; it cannot send marketing to someone who has not agreed to receive it; anything costly or unusual goes to a person first. Every decision — allowed, refused or escalated — is recorded, and the business can review it.
What it does not do
- It does not make decisions with legal or similarly significant effects about you on its own.
- It does not use your messages to train AI models.
- It does not share your messages with other businesses that use Neaion; each business's data is kept separate by the database itself.
Talking to a person
You can always ask to deal with a person. The business that answered you is responsible for your conversation and your data; contact it directly, and it will be able to see the full exchange. See the Privacy Policy for how your rights work.
In the United States
Some US states have their own rules on telling people they are dealing with AI — among them California's bot disclosure law (Cal. Bus. & Prof. Code § 17940 and following: a bot must not mislead anyone about being artificial to encourage a sale) and Utah's Artificial Intelligence Policy Act (Utah Code § 13-75-103: a business must say it is generative AI when someone clearly asks). Neaion does this in every place it speaks for a business:
- Email: every reply Neaion writes says, under the answer, that it was written with AI for the business, with a link to this page.
- The website chat says it is an AI on its button and at the top of the chat, before anyone writes.
- Telegram: the business's own bot says so in its first message, and replies in a chat say so at least once a day.
- When someone asks — "are you a bot?", "am I talking to a real person?" — on any channel, the answer is a fixed sentence written by people, not by the AI, in the asker's language: yes, it is an AI assistant working for the business, and a person on the team can take over. It never claims to be human and never dodges the question.
Some licensed professions have stricter duties for conversations a state treats as high-risk (for example health, financial or legal advice). Neaion is not built to give such advice: its practice playbook says no advice by message, and it is not for patients' health information (Terms, section 14).