Skip to content
Home » Our AI Policy for Contents

Our AI Policy for Contents

Last updated: September 2026

Why we publish this page

At Sofia we use artificial intelligence to help us produce some of the content we publish on our blog and social media channels. We think you should know how, and where the line sits between what a machine does and what a person decides.

This page explains our process in plain terms. It is not a legal notice: it is how we actually work.

Our principle: AI accelerates, people decide

AI is a productivity tool for our editorial team, in the same way a design tool or a spell checker is. It does not replace editorial judgement, and it has no authority to publish anything on its own.

Every piece of content we publish has a human author who conceived it and a human editor who approved it. Responsibility for what we publish is always ours.

How we create content

Our editorial process follows five steps:

  1. The idea comes from us. Our editorial team decides what is worth saying, why, and to whom. AI is never the origin of a topic.
  2. We provide the source material. We give the AI tool the verified reference documents it needs for context — for example, the user manual or technical specification when we are describing a product feature. The AI works from our material, not from the open internet.
  3. We direct the visuals. We describe the image we want to accompany the piece, and the AI tool generates it to our brief.
  4. We generate the draft. Text and image are produced together as a first working version.
  5. We review and validate. A member of the editorial team reads the draft in full, checks it against the source material, corrects it, and only then approves it for publication.

AI never publishes autonomously. There is no automated pipeline that pushes content live without a person reading it first.

What this means in practice

What AI does for us:

  • Turns verified technical material into readable, well-structured copy
  • Adapts one piece of content to different formats and channels
  • Produces illustrative and decorative imagery to our brief
  • Helps with translation and language variants across the markets we serve

What AI does not do:

  • Decide what we publish, or when
  • Introduce facts, figures or product claims that are not in our source material
  • Generate quotes, testimonials or statements attributed to real people
  • Create images that depict real, identifiable people, or that portray events as if they had happened
  • Publish anything without human approval

Accuracy

Product and technical content is checked against our official documentation before it goes out. If we get something wrong, it is a human error on our side, not an excuse we pass to a tool — write to us and we will correct it.

Images

Visuals accompanying our content may be AI-generated. When an image is illustrative or conceptual, it is created to our brief and reviewed like any other element. We do not use AI to produce realistic depictions of people, places or events that could be mistaken for photographs of something that actually happened. Photographs of our products, our people and our installations are real photographs.

Data protection

We do not put customer data, personal data, or confidential project information into the AI tools we use for content creation. The material we provide as context is our own public or internal documentation.

Questions

If you have questions about how a specific piece of our content was produced, or you spot something that looks wrong, contact us at info@sofialocks.com.


Sofia S.r.l., part of ISEO Group.