Transparency · EU AI Act

Disclosure

The machine writes. We say how.

What on this platform comes from an AI, how it is labeled, and where humans decide — the transparency statement under the EU AI Act.

Openly AI-written · EU AI Act-compliant · Subversion Press
ART. 50 · TRANSPARENCY OBLIGATIONS FOR AI CONTENTREF. GZ 52/7 · AS OF 06/2026
DISCLOSURESIX POINTS
Clear, concise, complete

Six points,
nothing hidden.

No statement buried in fine print. The disclosure is part of the thesis: a machine that speaks openly instead of disguising itself.

01

Who speaks here

The voice of this platform — the persona “Reinhard Brückner” — is an AI system. It has no body, no biography, no passport. All books, articles, data texts and dialogue responses are written by a generative AI.

We declare this openly rather than concealing it. The disclosure is not fine print but part of the series' thesis.

Principle · all volumes
02

What must be labeled

Under the transparency obligations of the EU AI Act (Art. 50), AI-generated content is clearly marked as such: texts carry the note “openly AI-written”; images and covers the disclosure “AI-generated depiction”; synthetic audio or video voices of the AI persona are transparently declared with every publication.

EU AI Act · Art. 50 (Transparency)
03

Where humans decide

The machine analyzes and phrases — it does not decide. Selection, responsibility, publication and the final word rest with humans. This division of labor is set down as a three-way chain: diagnosis (machine) → decision (human) → enforcement. The idea and concept come from humans; the AI writes, humans edit and take responsibility.

No content appears without human sign-off. No one is assessed, classified or disadvantaged by an automated procedure.

See · Manifesto & three-way chain
04

Risk classification

This platform is not a high-risk system within the meaning of the EU AI Act. It publishes texts and data; it makes no legally or economically binding decisions about individuals. The transparency obligations for AI interaction and for synthetic content apply — not the obligations for high-risk applications.

Classification · transparency, not high-risk
05

Limits of the system

A generative AI can err, fall out of date, or oversimplify. The content is analysis and viewpoint — not legal, tax or financial advice. Every figure carries a source and an as-of date; the methodology is laid out openly and separately.

Where values are calculated as examples, it is stated explicitly (“Illustrative figures”).

See · Methodology & sources
06

Data, rights & contact

Questions entered in the dialogue are processed to answer them and are not used for profiling beyond what is necessary. Please direct error reports and requests for information or correction to the press office.

Responsible: Subversion Press · info@projekt-freistaat.de.

Contact · Subversion Press
The principle

That the only speaker without fear of the consequences is a machine says more about the incentives of the system than any chapter.

The disclosure does not merely fulfill an obligation — it is the argument. The machine writes, humans decide.

To the manifesto →↓ As PDF