AI Disclosure
Pinchy shows a quiet AI assistant label in the chat header, right next to every agent's name. It is always present — on desktop and mobile — so a user opening a chat sees immediately that they are interacting with an AI system, not a human.
The label is deliberately unobtrusive: plain, muted text that reads as part of the agent's identity rather than a badge competing for attention. It carries no icon and no hover tooltip, because the disclosure has to be legible on every platform — including touch, where hover never happens.
EU AI Act Article 50 requires that people be informed when they interact with an AI system, at the latest at the time of the first interaction (unless it is already obvious from the context). A persistent, always-visible text label is the simplest, most durable form of that disclosure: it does not depend on a one-time notice the user might dismiss or scroll past, and it stays correct for every agent and every conversation.
What it says
Section titled “What it says”The label reads AI assistant. Every Pinchy agent is an AI assistant, so the wording is always accurate — there is nothing to configure per agent in v1.
Channels
Section titled “Channels”The in-app label covers the Pinchy web chat. Messages delivered through other channels are disclosed by that channel's own conventions:
- Telegram marks bots natively — a bot account's username ends in
botand Telegram displays a bot designation in the chat — so it is already obvious to the user that they are talking to an automated system.
When you connect a new channel that does not make the AI nature obvious on its own, add a disclosure to the agent's greeting or the channel's presentation so the Article 50 requirement is met there too.