Courtesy translation. This page is an English translation provided for convenience. The service is operated from Hungary and the legally binding text is the Hungarian original; in case of any discrepancy, the Hungarian version prevails.

We ask our clients to mark where a machine works on their behalf. So we publish ours. This page sets out where we use artificial intelligence, where a machine is talking to you, and who is answerable for the result.

Last updated: 26 August 2026. Legal background: Article 50 of the European Union's Artificial Intelligence Act, whose transparency rules apply from 2 August 2026, and Section 2:48 of the Hungarian Civil Code on the protection of image and voice recordings.

1. Where does a machine talk to you?

AI Control Alexandra is an AI-based sales assistant, not a human. She takes general enquiries, answers common questions and records details. Every message she sends carries a note at the bottom saying it was written by an AI assistant.

AI Control Vanda is an AI-based executive assistant, not a human. She supports internal work: calendar, correspondence, task management, summaries. She does not correspond with clients directly.

The human you can always ask for: Alex Kis, the founder. If you want to speak to a person, say so and a person answers. That is not an add-on, it is the default.

Chatbot: there is none on this website at the moment. If we add one, it will state clearly that you are talking to a machine before the conversation starts.

2. What we use AI for (itemised)

• Articles and blog posts: the outline and the draft are produced with AI support. Every published article is reviewed, checked and approved by a human. Editorial responsibility rests with Alex Kis.
• Social media posts: drafts with AI support, approved by a human.
• Videos (TikTok, Reels, Shorts): the narration is a synthetic voice and a significant part of the visuals is AI-generated. We state this in the videos and with the platform's own AI label.
• Illustrations, backgrounds and avatars on this website: partly AI-generated images. They do not depict real people.
• First pass over incoming enquiries: classification, summarising and draft answers with AI. The actual offer and the decision are human.
• Internal operations: calendar, administration, e-mail filtering, task management, summaries.
• Client projects: in the systems we build (automation, chatbots, content production) we record in writing with the client where AI works, and we hand over the labelling obligations on delivery.

3. What we do not do

• We do not pass a machine's work off as a human's. If AI answers, we say so.
• We do not produce unlabelled deepfakes: AI-generated or modified images, audio or video of a real person, place or event are always labelled here.
• We do not use anyone's face or voice without written consent. Voice cloning, face swapping and generated avatars only happen with the permission of the person concerned.
• We do not use emotion recognition or biometric categorisation systems.
• We do not hand client data to an AI provider for model training.

4. Who is answerable for the result?

A human. Every deliverable, offer and published piece is approved by a person. AI makes the work faster; it does not make the decisions. Responsibility on behalf of the business rests with Alex Kis, sole trader, whose details are in the Imprint.

5. Data processing

What data we process, for how long and for what purpose is set out in the Privacy Policy. Using AI tools does not change the legal basis or the purposes of the processing.

6. Comments and complaints

If you think any of our content or systems is not labelled properly, write to info@aicm.pro. We will fix it and tell you what we changed. The supervisory authority for the AI Act in Hungary is the National Media and Infocommunications Authority (NMHH).

This page is information, not legal advice. We update the statement when the legal environment changes.