Legal

AI labelling at NexAI

We use AI systems, for example a chat assistant, a phone assistant and an email assistant, and we also operate them for our customers. We label content generated by these systems as artificially generated (Article 50 of the EU AI Act). This page explains how to recognise it and how to check it yourself.

Emails. Replies drafted by our email assistant carry two header fields:

X-AI-Generated: yes
X-AI-Disclosure, which names the system that generated the reply and states whether a person reviewed it before sending (review=human) or not (review=none).

To view the message header:

in Gmail use "Show original",
in Outlook go to "File" → "Properties" → "Internet headers",
in Apple Mail go to "View" → "Message" → "All Headers".

Chat. In the chat window, the first message states that an AI system is replying, and a visible "AI" label stays in the window while it is open. Each reply from the assistant is also labelled as AI-generated in the page source (data-ai-generated="true"). Replies are kept deliberately short.

Phone. At the start of every call, the assistant announces that you are speaking with an AI assistant. It announces any transfer to a person beforehand. Calls are not recorded as audio.

Questions. If you want to know whether a specific piece of content came from one of our systems, write to us at mbt@nex-a-i.com.