AI transparency

We use AI models where they make us better and faster:

• Text: research, structure and drafts for blog posts and website copy
• Translation: first drafts, for example of the English website texts
• Programming: code assistance in development, this website included
• Images and video: editing and – where labelled – creating graphics

For all of it: a person on our team reviews, edits and approves the result and takes responsibility for it. Nothing goes out unchecked.

We would rather label once too often than once too rarely:

• All texts on this website – pages and blog posts alike – were created with AI assistance and reviewed and approved by a person on our team. The footer of every page says so, too.
• Blog posts also carry "Created with AI assistance, reviewed by …" naming the responsible person, or "Created with AI" when the text mainly comes from an AI.
• Translated testimonials are marked as translations; the original remains authoritative.
• AI-generated images are labelled in the caption and keep their machine-readable provenance data. Photos of our team and our projects are real.
• Every graphic we generated with AI carries "AI-generated" or "Grafik KI-generiert" (AI-generated graphic) in the image itself: blog thumbnails, the illustrations of our services, the previews in the configurator, app covers and the share images of every page.
• Project covers are graphics designed with AI around real screenshots of our projects and carry the label "Grafik KI-generiert" (AI-generated graphic). We built the projects themselves; the screenshots and photos in them are real.
• The home page video and the video banners are animations programmed with AI, partly with licensed stock footage (Pexels), and labelled "AI-generated". The other videos are real footage by Pexels photographers.

The German texts are the originals. The English texts may be AI-assisted translations that we have read, adapted and approved. For legal texts the German version always prevails.

There is no AI chatbot and no voice assistant on this website. When you write or call, people answer – usually Christoph or Lukas in person.

People read and answer inquiries sent via the contact form or the configurator. We do not rate or sort inquiries with AI, and we do not pass data you enter here to AI services. The spam protection uses fixed rules and a computing puzzle, not AI (see the privacy policy).

We build AI solutions and automation for businesses – from assistants for internal processes to document processing. Labelling, human oversight and data protection are part of the plan from day one. We only pass confidential client data to AI providers with the client's consent and under a data processing agreement. Learn more under AI & automation.

Everyone at Proplex who works with AI systems is trained in the tools we use, their limits and risks (Art. 4 AI Act). We are happy to answer questions about our use of AI at office@proplex.io.