Lotus LibreAI systems
Book audit

AI transparency statement

Last updated:

We build AI systems for a living, so we hold ourselves to the transparency we ask of others. This statement explains where AI is used on this website, in our content and in the systems we deliver, and how we meet Article 50 of the EU Artificial Intelligence Act (Regulation (EU) 2024/1689).

On this website

  • The interactive demos are simulations. The agent log, the voice call, the chat, the creative lab, the build sequence and the gateway dashboard on our pages are scripted animations running in your browser. They do not call any AI model, do not record you and do not send data anywhere. Figures shown in them, and figures marked “example”, illustrate typical outcomes; they are not live data.
  • No chatbot, no profiling. There is no AI assistant on this website, and we do not analyse your behaviour, because we don’t track it at all.
  • Images. Illustrations on this site are drawn in code. We do not use photographs of real people.

In our content

Our articles and their translations are researched and drafted with the help of AI models and then edited and published under the editorial responsibility of Lotus Libre. Every factual claim is meant to be backed by a cited source; if you find an error, tell us at hello@lotuslibre.com and we will correct it.

Your requests

Requests sent through our contact form are read by people. We do not make automated decisions about you within the meaning of Article 22 GDPR.

In the systems we deliver

Article 50 of the EU AI Act applies from 2 August 2026. For every client system we build, we design for it by default:

  1. Disclosure in conversations. Chat and voice agents tell people they are interacting with an AI system at the first interaction, unless this is obvious from the context.
  2. Labelling of synthetic content. Realistic AI-generated or manipulated images, audio and video (“deep fakes”) are disclosed with visible labels where required, and we add machine-readable provenance metadata (C2PA) by default.
  3. No prohibited practices. We do not build emotion recognition for workplaces or education, which has been prohibited in the EU since 2 February 2025, and we never clone a real person’s voice or likeness without their documented consent.
  4. Human oversight. Risky or irreversible actions — refunds above a limit, legal or medical topics, anything a client defines — are routed to a person.
  5. Logs and evaluation. Agent actions are logged and evaluated so that clients can explain what their systems did.

For systems that may fall into a high-risk category under Annex III of the AI Act (for example recruitment or credit scoring), we discuss the obligations that apply from 2 December 2027 before any work starts — and decline projects we cannot deliver responsibly.

Questions

Write to hello@lotuslibre.com. We are happy to explain any of this in more detail.