AI disclosure

We use AI as a tool in producing content and imagery for this site. On this page we explain openly where and how — not because the law would require it for everything, but because we consider it the right way to work.

Texts

AI has been used in producing the texts on this site. All published text goes through human review before publication: the content is read, the facts are checked and a human decides what is published. AI never publishes anything on its own.

Images

Photorealistic images produced entirely or substantially with AI are marked with a visible AI label in the image itself, so the label travels with the image when it is shared onwards.

The label is not used for content that does not depict reality photographically:

  • illustrations and drawings
  • icons, charts and infographics
  • device mockups and other clearly staged presentation images

The site may also contain ordinary photographs, the site owner's own images and stock photos. The absence of a label is therefore not a guarantee of an image's origin — it tells you the image is not a photorealistic AI image.

Chat assistant

The chat on this site is answered by an AI assistant. This is stated at the start of the conversation and shown in the chat window throughout. The assistant can make mistakes — please confirm important matters with us directly.

Conversations may be reviewed to ensure service quality. Please do not write sensitive or otherwise confidential information into the chat beyond what using the service requires. The processing of personal data is described in our privacy policy.

Responsibility for content

Responsibility for the content of this site lies with NOX Visual Oy. Using AI as a tool does not reduce that responsibility. If you notice an error, something unclear, or content you believe should be labelled differently, let us know: info@noxvisual.fi. We correct identified errors without delay.

What this is based on

The EU Artificial Intelligence Act (2024/1689) requires transparency for certain AI-generated content and for interacting with AI. Our practice is deliberately broader than the legal minimum: we label photorealistic AI images even when labelling would not be mandatory, and we do not interpret borderline cases in our own favour.