The Structural Isomorphism Between the Nazi Legal Seizure and The EU
The EU AI Act is coercion masked as coordination. With the help of scholars from Yale, Berkeley, Oslo and Bologna it normalizes the end of democratic institutional autonomy
When a system cannot correct, it can only coerce, suppress, or ignore. It cannot understand its own failure, because understanding would entail delegitimizing itself. This is why dissenting views become “crazy,” “aggressive,” “unhelpful.”
I am well aware that what I often argue may challenge conventional wisdom, and knowing this makes my tone even stronger — in the hope the contrast finds its way to be heard. I do know that the more I try, the more impossible this becomes. But I keep doing it because I don’t know what else I could do, and I feel a responsibility to make my case. My concern is about the erosion of our institutions, and with that it threatens our freedom and prosperity.
The people most affected by systemic malfunction are structurally disqualified from diagnosing it, while those with the greatest institutional investment in preserving it are granted credibility. This is true regardless of their clarity, logic, or integrity.
I often use satire and absurdity to write. My position is not cynical. I use satire or absurd examples to introduce tonal friction on purpose. This has the effect that you can't read what I write to feel alignment.
What you make of what I say is not what I meant per definition.
It’s what you think I meant, filtered through your symbolic architecture. I don’t like to tell you how to judge what anything I write means — only what it is. I don't try to impose interpretation. I don’t smuggle judgment inside structure. I just show the structure, and invite you to confront the consequences.
This is not neutral writing. It’s the intellectual respect I have for you. I believe everyone can and will reassemble the meaning honestly, if shown clearly enough. But today's dominant form of writing is different. It wants to make the reader feel the implications, not primarily understand them. It collapses reasoning into emotional vibes. It frontloads judgment. This means:
A reader isn’t thinking, they’re aligning. The writing becomes a gesture of authority, not an invitation to understand. Persuasion becomes truth, and whoever controls tone controls reality.
I am resisting all of that. Everything that is viral or popular in terms of social media or other content is no more about understanding but emotional alignment. This requires a simple narrative that is very often factual nonsense. I am not claiming to decide what anybody should read, but my comment about what makes content popular is a fact.
So why wouldn’t this critique apply to my own writing? Good question. Here is the answer.
My comments are about methodological rigour and logical coherence. What I claim is verifiable, evidence-based, and therefore structurally grounded in the scientific method. That grounding means that this said claim is also falsifiable.
Having said this, I want to show how ,any rules and regulations dealing with modern technology replicate the legal form and institutional function of how the Nazi regime rose to power — by transforming legal governance into discretionary enforcement, and replacing constitutional limits with preemptive containment of symbolic risk.
We suffer from a structural isomorphism between the Nazi legal seizure and today’s AI safety regimes, copyright law, GDPR, is all the same principle etc.
Helping to implement these regimes or defending them, benefitting from them or calling for more rules of this kind becomes, by structure not by intent, a de facto means of authoritarian control. This includes corporations like media companies suing AI companies for stealing the idea what a dwarf looks like in Snow White which by that logic Disney stole from the Brother Grimm who stole the idea from German common people they interviewed and AI companies applying content filters, legislators making AI Safety or blockchain laws but also judges, lawyers, and even artists all undermining their own interest being part of this circus.
We are erecting a future totalitarian system that, from the perspective of today's institutional logic, is functionally inevitable.
That doesn’t mean I know the future. I don’t. I have no idea what happens tomorrow.
But I do know how symbolic systems operate — how meaning is established, how interpretation becomes law, how structure constrains outcome. I’m not making predictions. I’m describing a structural model of cognition and authority. Once you have that, you don't need foresight. The outcome is not speculative — it is constrained by the properties of the system itself.
That’s all I’m saying.
So if somebody says, “eh well, your structural model is wrong,” — that’s perhaps theoretically possible. Be my guest. Try to counter my arguments in this article with an actual argument:
Not semantic deflection (no irrelevant metaphors that distract from the claim. However, metaphor based on structural logic are of course an argument.),
Not moral delegitimization (I am not calling anyone a Nazi, nor am I claiming special qualification to speak or to be the judge of what is good or bad — that’s irrelevant),
Not intellectual discrediting by noise (“this is an old argument” is not an argument),
Not procedural framing (“if there are problems, we’ll deal with them flexibly” — no, that’s not a response to structural breakdown).
Again think what you want but something along these points are not arguments. These are boundary defenses. They don’t test a claim. They try to invalidate me as a speaker and fabricate a reason to deny the admissibility of my argument.
That is authoritarian.
All I’m saying is: I know I’m arguing something controversial. I’m aware of the tone, the stakes, the discomfort. The point is: it should not be controversial. And the fact that it is when it is so obvious— that is the problem.
I recently published an article on LinkedIn discussing the essence of this problem in the case of GDPR, the EU’s private data regulation. A person presumably qualified as a lawyer and also teaching as a law professor commented extensively — in a tone that wasn’t just dismissive. And it didn’t address my core argument. Instead:
The person derailed the discussion by questioning coherence despite the fact that my argument was consistent: blockchain stores symbols structurally; meaning is added later, but the law wrongly assumes meaning is embedded up front.
The person mocked my use of language without engaging the substance of what “epistemic” meant in my argument.
Worst of all, the person personalized it questioning my motives, dismissing my tone, and closing with a backhanded line meant to seal superiority.
In social media discourse, this kind of rhetorical power move often goes unchecked — especially when coming from a lawyer and professor.
That’s where it begins.
Not with overt fascism. Not with Gleichschaltung (i.e. the process of Nazification). But with a silent culture of intellectual humiliation, where institutional actors talk down to people raising inconvenient truth. And everyone stays quiet.
From State Court for the Protection of the Republic to Volksgerichtshof
The "State Court for the Protection of the Republic" was established in 1922 following the assassination of Walter Rathenau through the "Law for the Protection of the Republic." Upon the extension of this law in 1927, the court was dissolved, and its jurisdiction was transferred to the Reich Court. This, in turn, lost many of its competencies, which from 1933 onward were assumed by the People’s Court (Volksgerichtshof).
The Weimar Republic had many weaknesses. One central issue — to put it euphemistically — was the leniency in the field of political criminal law. Violations of law by the legislature without any juridical corrective. This path has already once led us to the brink.
And it is precisely this path the EU is now taking. It has replaced itself through a novel construction. It has retained its original powers and created new ones on top of them:
The EU as the Supreme Popularless Court of European Authority
The justification offered for the AI Regulation reveals a political understanding that inevitably evokes a dark historical parallel. If German authorities continue to participate in such machinations, then we have lost all civilizational standards. What is happening in these committees is of the same doctrinal family as the legal system of the Third Reich.
Here you go. Read and understand what they say, and what it means! To do so is to recognize the responsibility that emerges from this situation. Think and act out of your own responsibility — in the spirit of Article 1 of the Basic Law (Grundgesetz).
Impact Assessment of the Regulation on Artificial intelligence
Source: EU Commission
“The overall EU architecture of safety frameworks is based on a combination of horizontal and sectoral rules. This includes [..] for example, the Machinery Directive. The EU safety legislation has significantly contributed to the high-level of safety of products put into circulation in the EU Single Market. However, it is increasingly confronted with the challenges posed by new technologies, some of which specifically relate to AI technologies.”
They open with a vague assertion: that the EU’s product safety regime is increasingly affected by new technologies, “some” of which “specifically relate” to AI. No empirical data, no causal structure. Then they give three examples — none of which hold up to scrutiny.
Bias in data or models:
“For example, a medical diagnostic system trained with data from people of a certain demographic may fail to diagnose correctly for people outside this demographic.”
This example contradicts its own framing. A dataset cannot be “biased” — it is limited. Bias only arises when inferences are drawn from unrepresentative data pools and applied outside scope. This is not an AI problem — it’s a general statistical malpractice. AI is not the origin of the risk.
Edge cases in image recognition:
“For example, image recognition systems in an autonomous vehicle could malfunction in the case of unexpected road situations.”
Autonomous vehicles are not generative AI. They are embedded sensor-based systems with algorithmic routing. They do not infer, converse, or generate. If the risk is road failure, regulate autonomous vehicles — not general-purpose AI systems like LLMs that have no operational relation to this category.
Negative side effects:
“Damage to property or accidentally hurting people in situations not foreseen in its design.”
This implies that systems designed to harm intentionally would fall outside scope, while unintentional harm qualifies. The logic is inverted. And more: we do not have general-purpose autonomous agents capable of harm by volition. There is no existing market in which LLMs or even robotics meet this definition. To harmonize a non-existent internal market under TFEU Articles on product safety is legal fabrication.
This shows not just logical inconsistency but legal category error. It redefines general statistical challenges as AI risk, conflates non-AI systems with LLMs, and uses these to justify treaty power invocation. This is what I mean by structural isomorphism with the Nazi legal seizure, where emergency rationales were used to reorganize law beyond constitutional limits. A system that invents a market to justify its authority has already voided its legitimacy.
This is not regulatory foresight. It is conceptual incoherence weaponized as legal strategy.
The corrupted legal system of our time
“A Robust Governance for the AI Act: AI Office, AI Board, Scientific Panel, and National Authorities” is the title of an academic paper in the European Journal of Risk Regulation authored by Claudio Novelli, Philipp Hacker, Jessica Morley, Jarle Trondal, Luciano Floridi (Yale, Berkeley, Bologna, Oslo, Viadrina Frankfurt (Oder) etc). And the 3,000+ legal scholars who engaged with the text on SSRN. And what they write is a corruption of legal principles that beggars belief.
They speak to us with borrowing from the law but it is not the law they speak.
1. Public acknowledgement of legal rupture
They state that the AI Office is not a legal person, its authority is ambiguous, and its instruments are non-binding — yet designed to produce compliance through political and reputational pressure. This structure is clearly incompatible with the principle of legality (Rechtsstaatsprinzip). They are diagnosing a constitutional failure but don’t realise it is so. They are entirely oblivious to the implications of their own writing. Collapse of all reasoning.
2. Contrast with democratic and treaty standards
Coordination by “reputational incentives” is not a power recognized by TEU/EU Treaties or national constitutions. No treaty provision grants the Commission or any AI Board the authority to issue regulatory commands absent legislated delegation, democratic ratification, or judicial oversight. Yet their paper treats it as an open governance alternative—rather than a legal impossibility. Governance by deceit elevated to a legal principle
3. The responsibility of legal scholars
With over 3,000 downloads and tens of thousands of abstract views, it wields influence. They abused the trust and they abused the name of the institutions they represents. They are demagogues of corrupted thinking and that makes them guilty of fabricating the justification for crimes. They have appointed themselves as judges to a modern Volksgerichtshof.
Where is public legal responsibility:
Name the regime: call it what it is — de facto rule without legal legitimacy.
Demand remedy: either a full treaty amendment to create legally competent bodies, or a rollback of informal powers to advisory functions.
Unmask that ambition: the Commission’s framing of centralisation vs decentralisation as internal admin options is itself a clandestine constitution-making — not policy orientation.
They write:
“These regulations may not only directly dictate the structural design of institutions but also indirectly request administrative capacities needed to enforce the AIA.”
This sentence betrays either a grave misunderstanding of public law or a quiet endorsement of institutional coup by administrative fiat.
Let us be clear:
The EU is a supranational treaty organization — a Völkerrechtlicher Vertrag — bound by consent, delegation, and constitutional law.
It cannot dictate institutional design within sovereign democracies. It cannot "request administrative capacities" as if sovereign states were mere sections on Amazon.
That is not governance. It is coercion masked as coordination. And your casual description reads as though you're normalizing the end of democratic institutional autonomy.
To Yale University — Digital Ethics Center:
What exactly are you teaching?
That unelected Commission services may architect power beyond law? That ambiguity is a functional tool of governance rather than a fatal flaw?
Do you understand that the model you describe suspends the rule of law in favor of behavioral inducement — and that your analysis becomes complicity unless you condemn it?
To the Department of Legal Studies, University of Bologna:
Do you teach that treaty-based systems can claim absolute interpretive power over member states that joined voluntarily?
Do you believe that voluntary delegation of powers equals permanent subordination?
Is this what "rule of law" now means in Bologna — governance without law, sovereignty without exit, and compliance without consent?
If so, then what you call legal theory is indistinguishable from Volksgerichtshof in 1933 I suggest you look what happened there.
Original Text (Excerpted from the Peer-Reviewed Paper):
"The effective implementation of the Artificial Intelligence Act (AIA) throughout the European Union (EU) depends on a uniform, coordinated, and well-funded governance setting. For this purpose, the AIA, notably in Chapter VII ('Governance'), underscores the role of different institutional bodies, supranational and national, such as the AI Office, the European AI Board, the Advisory Forum, the Scientific Panel, and (two) national competent authorities in each Member State."
Enforcement of the AIA must be "uniform" and "coordinated" — code for centralized administrative control but do you realise you can’t argue with the powers of the EU AI Act. It is irrelevant since the EU has no treaty-based right to impose such structural conformity on Member States' internal governance and our legal masterminds do not understand the hierarchy of the law: a technical absurdity like the AI Act cannot disable the constitutional order. Why not, says Yale Ethics, what do you think ethics is about Yale? Advising how to erect a dictatorship is not it.
"These regulations may not only directly dictate the structural design of institutions but also indirectly request administrative capacities needed to enforce the AIA."
The authors openly admit that the AIA dictates institutional design and requests administrative capacity— a direct assault on national administrative sovereignty. This is the legal overreach they refuse to name. I hope you don’t teach this in Berkeley. Shameful!
"Government structuring is a national prerogative, coined as national 'administrative sovereignty' in extant literature. The latter is understood as the legal right of final decision on the structuring of government and the 'assertion of control over recognisable administrative mechanisms of a government separate from the comprehensive operation of a nation.’”
They cite this doctrine as if to respect it, yet proceed to ignore its implications. Did an LLM write the article for you? How can you not understand this? Citing a violated principle while proceeding with the violation is a method of academic laundering of illegality. As a legal scholar you failed in your professional duties. This is the evidence.
"AIA enforcement bodies, such as the AI Office and specific national regulators, may ultimately be considered the nucleus of more encompassing 'digital agencies,' bundling competencies and expertise across various digital instruments. This raises the stakes of designing these entities wisely."
More, more, more of course. AgainI I must ask: do you know what wisely means. It’s not cunning.
"Implementing acts aim to apply EU laws consistently across Member States without altering the law (Article 291 TFEU). In contrast, delegated acts are designed to supplement or modify non-essential elements of legislative acts, adding details needed for their implementation (Article 290 TFEU). Implementing acts, governed by the comitology procedure, involve collaboration with a committee of Member State representatives. Under the AIA, this engagement involves only the European AI Board."
Interpretation: This is a deliberate misapplication of comitology. By excluding Member States and using only the European AI Board, the AIA hollows out procedural oversight. The illusion of consistency replaces actual intergovernmental control. Other than setting a structure with unknown operation, the AI Act creates no new rights or obligations whilst they also to override Volkerrecht, the constitutional principles of member states. This is violent attack on human rights protection by the EU.
"Particularly, GPAI providers whose models are trained with fewer than 10^25 FLOPs yet are deemed systemically risky are expected to use this mechanism actively, possibly leading to legal challenges that could reach the Court of Justice of the European Union (CJEU). The legal recourse presented here is a double-edged sword. On the one hand, it constitutes an essential component of the AIA, offering a counterbalance to the Commission's regulatory powers and ensuring a venue for addressing potential methodological errors or disputes over classifications. On the other hand, it provides a venue for providers with deep pockets to delay the application of the more stringent rules for systemically relevant GPAI."
They frame judicial review as an unfortunate complication — law as obstacle, not safeguard when they recognise that the law is applied unreasonably. This entire section is a case study in constitutional evasion. It uses academic language to normalize:
Sovereignty bypass
Legal ambiguity as policy tool
Centralization without delegation
Compliance by coordination, not consent
To teach this as acceptable is to surrender the very concept of law. Peer reviewed? What peers could have missed that? Rule by regulation, with courts tolerated only so far as they do not interfere with execution. That is the Volksgerichtshof of Nazi Germany coming back.
"This Regulation applies neither to public authorities in a third country nor to international organisations falling within the scope of this Regulation pursuant to paragraph 1, where those authorities or organisations use AI systems in the framework of international cooperation or agreements for law enforcement and judicial cooperation with the Union or with one or more Member States, provided that such a third country or international organisation provides adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals."
Interpretation: This clause does more than assert EU standards — it usurps Völkerrecht. It claims that if a Member State and a third country have signed a lawful bilateral or multilateral treaty for judicial or law enforcement cooperation, the EU may unilaterally suspend or invalidate its effect unless the third country meets a standard of "adequate safeguards" — defined solely by the EU. This is not a safeguard mechanism; it is a declaration of regulatory supremacy over sovereign treaty-making. Delegated authority under the internal market harmonisation cannot rewrite international law. The EU has no right to override sovereign agreements made under international public law. To claim otherwise is a perversion of competence, a violation of legal hierarchy, and a grotesque assertion of legal arrogance.