How we use AI
We build with AI every day. This page states where, what it means for the work we deliver, and which parts a human is accountable for. It also covers our duties under Article 50 of the EU AI Act, which applies from 2 August 2026.
Last updated on 2 August 2026.
The short version
AI helps us write, design, build and analyse. A human decides, checks and carries responsibility for everything we publish or hand over. Where an image could pass for a real photograph but is not one, we say so next to the image.
Text on this site and in our work
Blog posts, page copy, emails and proposals are drafted with AI support and then edited by a person who checks the facts, the sources and the tone. That person carries editorial responsibility. Under Article 50(4) of the AI Act no label is required in that case, and we do not add one, because a label on everything makes the label meaningless.
Statistics and legal statements are traced back to their primary source and linked. If we cannot find the source, we say so instead of quoting the number anyway.
Images and video
Our diagrams, charts and illustrations are drawn by us with AI assistance. They are clearly graphic rather than photographic, so they are not deepfakes and carry no label.
Photorealistic images that were generated or substantially altered by AI do carry a visible note next to the image, in the caption or the credit line. That includes AI-generated portrait photography of our own team.
Colour correction, lighting, noise reduction, upscaling and light cropping with AI do not change what an image shows, and are not labelled.
Chat and voice assistants
Our chat and voice assistants identify themselves as AI at the start of the conversation. A human takes over as soon as the assistant cannot help, and always for anything involving a quote, a complaint or a payment.
Code and systems
We write software with AI programming assistants. The AI Act regulates what software does, not how it was written, so there is no labelling duty here. We do check that generated code does not infringe third-party licences, and we remain responsible for security and correctness.
Your data
We do not feed client data into public AI tools for model training. Where AI processes personal data on our behalf, that processing falls under our privacy policy and the agreements we hold with our suppliers.
Found something we got wrong?
If you spot content that should carry a label and does not, tell us at info@teammade.ai. We will correct it and note the date on this page.