When Germany moralizes away its own future – and why that is an economic, cultural and social failure

Xpert Pre-Release

Language selection 📢

Published on: June 11, 2026 / Updated on: June 11, 2026 – Author: Konrad Wolfenstein

When Germany moralizes away its own future – and why that is an economic, cultural and social failure – Image: Xpert.Digital

The FAZ scandal that wasn't: How a faulty AI detector sparked a national debate

Technological fear as a virtue: How Germany is moralizing away its own future

Deleted due to 100% “AI suspicion”: The Mario Voigt case illustrates the entire German digital dilemma

A deleted guest post, an unreliable algorithm, and a media frenzy that completely misses the point: The Frankfurter Allgemeine Zeitung's decision to remove an article by Thuringian Minister-President Mario Voigt due to a supposed "AI suspicion" is far more than a mere editorial footnote. The incident is a symptom of a German malaise. While the rest of the world has long been pragmatically using generative artificial intelligence to increase productivity and inclusion, Germany celebrates technological skepticism as moral superiority. Instead of discussing urgently needed youth protection measures, the public is lost in a hysteria surrounding tools that have long been part of everyday work life. This is an in-depth analysis of faulty software, the fatal economics of media outrage, and a country that risks simply moralizing away its economic and social future.

World market leader in slowing down: The FAZ-Voigt incident reflects a deeper problem

Outrage instead of facts: What the removal of a FAZ article reveals about our debate culture

On June 10, 2026, a guest article disappeared from the digital archive of the Frankfurter Allgemeine Zeitung (FAZ). The author was Mario Voigt, Minister-President of Thuringia. The title was "Smartphone 14, Social Media 16," and the publication date was August 13, 2025. The reason for its removal: suspected AI. The FAZ had the text analyzed by the AI detector Pangram and received a result indicating an alleged 100 percent AI content. Furthermore, three direct quotes—attributed to psychologist Jonathan Haidt, neurobiologist Gerald Hüther, and neuroscientist Manfred Spitzer—could not be verified. The editorial team decided: The article would be deleted.

What appears to be a routine editorial action is in reality a symptomatic event. It encapsulates in a single instance what has been going wrong in Germany for years: a culture of debate that celebrates technological skepticism as a virtue, engages in moralizing instead of analysis, and overlooks the fact that the rest of the world has long since moved on. This article analyzes the case of Mario Voigt and the FAZ (Frankfurter Allgemeine Zeitung) as an exemplary starting point for a thorough assessment – economic, social, and political.

What actually happened: Facts without hysteria

In the deleted post, Voigt called for a clear protection program for children in the digital sphere: smartphones only from the age of 14, social media only from 16, and a general ban on smartphones in primary school. He cited studies showing that one in four children suffers from anxiety due to social media, and findings on depressive symptoms in adolescents caused by excessive social media use. These are not fringe positions. Cem Özdemir of the Green Party publicly made the same demand. Voigt later reaffirmed his stance in the Thuringian state parliament, pointing out that mental illness among children has doubled in recent years.

The article's content was therefore at least legitimately debatable – and highly relevant to society. However, this was hardly discussed after it was taken down. From then on, the public debated how the text came about, not its content. This is telling.

The FAZ itself admitted that Pangram was "by no means perfect" and did not provide definitive proof. Nevertheless, it made a final decision. Voigt's state chancellery responded to the editorial team's inquiry, stating that AI would be "part of the everyday work of modern organizations by 2026," and that responsibility would always remain with humans. This answer was not sufficient for the FAZ. The article disappeared.

There was little new about this: It had already become known in early June 2026 that Voigt, together with the Minister-President of Saxony-Anhalt, Sven Schulze, had commissioned an AI-assisted guest article for Die Welt. The topic: more German-language music on the radio. Voigt's state chancellery confirmed at the time that they used "modern digital tools, including AI applications," but that the authors were responsible for the content. Thuringia's Digital Minister, Steffen Schütz, also advocated in this context for mandatory labeling of AI-generated texts.

What remains is a case that extends far beyond Voigt and the FAZ. Because it is not an isolated case – it is a pattern.

The technology in question: A sober assessment

AI-powered text production is a reality today. It's not a scandal; it's a tool—like a calculator, word processor, or search engine. According to the Federal Statistical Office, by 2025, 26 percent of all companies in Germany with at least ten employees were already using AI technologies, an increase of 14 percentage points compared to 2023. Among large companies with 250 or more employees, the usage rate was 57 percent. Generative AI—that is, the form of AI that produces texts, images, and content—was already in use in 18 percent of German companies by 2025, although this figure was close to zero in 2023.

According to a KPMG study from 2025, 91 percent of German companies now see generative AI as an important topic for their business model and future value creation, and 82 percent plan to increase their AI budgets in the next twelve months. This is no longer a fringe phenomenon. It's mainstream – and specifically, economic mainstream.

IBM demonstrated in a comprehensive study of 3,500 executives in ten countries that two-thirds of German companies have already achieved significant productivity gains through the use of AI. Approximately one in five companies in Germany has already met its ROI targets through AI-driven initiatives. The figures are clear: AI use has long since moved from a niche market to the mainstream. Anyone who fundamentally questions AI assistance in text production is questioning the reality of working life in 2026.

The measurement problem: When the detector judges

A key aspect of the Voigt case, which was largely overshadowed by the media outrage, is the questionable reliability of the measuring instrument used. The AI detector Pangram delivered a result of 100 percent AI content – thus triggering the entire debate. But how reliable is this assessment?

Scientific analyses show that Pangram had a false positive rate of two percent in studies by the University of Maryland and Microsoft. That sounds small, but it isn't. In a university setting with thousands of texts, this means that, statistically speaking, a significant proportion of human-written texts are incorrectly classified as AI-generated. The Higher Education Forum on Digitalization has also pointed out the systematic bias effects in AI detectors: Texts by people writing German as a second language, those who use particularly clear or structured language, or those who formulate according to a specific pattern are disproportionately often flagged as AI-generated.

The FAZ itself admitted that Pangram provided "no conclusive proof." Nevertheless, it made a final decision based on this imperfect evidence. This is a journalistic practice that is difficult to reconcile with its own claim to thoroughness.

The fundamental problem is epistemic: style is not proof. A well-written, structured, clear text—that is, a text that is technically convincing—is more frequently classified as AI-generated by AI detectors than a poorly formulated, contradictory text. This creates a perverse incentive: those who write clearly are under suspicion, while those who write awkwardly are considered authentically human.

The inclusion dimension: Who pays the price of this morality?

There is a group of people for whom this discourse is particularly consequential and who are hardly represented in the public debate: people with physical or cognitive impairments who rely on AI tools to be able to express themselves adequately.

AI has an emancipatory dimension for people with disabilities that can hardly be overestimated. Automatic speech recognition, real-time translation, text assistance, and formulation aids help people with hearing impairments, motor limitations, dyscalculia, dyslexia, or other disabilities to participate fully in a world dominated by written language. AI can break down learning barriers, strengthen self-determination, and promote social inclusion. For many of these people, AI assistance is not a convenience tool—it is a fundamental prerequisite for equal communication.

When a debate distorts who uses AI, as if that in itself were suspicious, it hits first and hardest those who have no choice. They can't simply forgo AI assistance and write "authentically human." If their texts are run through AI detectors, they may be flagged and delegitimized—not because they lied, but because they're using a tool they need. Equating AI use with dishonesty is therefore not only analytically imprecise; it is profoundly ableist.

The self-appointed moral guardian: Analysis of a phenomenon

Who was the first to raise concerns about the Voigt article? The online portal "Frag den Staat" (Ask the State) ran the text through Pangram and published the results. Journalist Jonathan Peaceman had previously drawn attention to the Welt article on the Bluesky network. This sparked a wave of coverage – Tagesspiegel, Bild, t-online, and even the FAZ itself.

The pattern is familiar and always follows the same pattern: Someone with a large following throws a vague accusation into the digital sphere, other media outlets pick it up, the accusation takes on a life of its own, and the accused has to defend themselves. Whether the original claim is valid becomes irrelevant. What counts is the response.

What's missing from this mechanism is what Johannes Volkmann – Helmut Kohl's grandson and a young CDU politician – identified on Markus Lanz's talk show: substance. Volkmann aptly criticized the fact that political talk shows and media discourse in general focus primarily on emotions and "not a single substantive issue" currently facing the country. It's about the feeling of outrage, not about analyzing the problem.

The media authorities themselves have pointed to this situation. At their 2025 annual conference, Chairwoman Dr. Eva Flecken demanded: "We must get out of the outrage centrifuge and into a culture of debate that has substance – not just clicks." This is a remarkably self-critical admission. At the same time, it shows that the problem is systemic: clicks and outrage are economic incentives that drive the business model of many media outlets. Moral outrage sells. Sober analysis often doesn't.

The long-term study on media trust from 2024 already showed that the perceived coarsening of public discourse in Germany is at an all-time high and negatively correlated with trust in the media and politics. At the same time, media cynicism is growing: the proportion of people who believe that the media in Germany undermines freedom of expression is increasing. One has to ask whether actions such as the removal of a political guest article based on algorithmic suspicion strengthen or weaken this trust.

Germany's structural AI dilemma: World market leader in the brakes

Behind the Voigt case lies a structural problem that is increasingly isolating Germany internationally. While 73 percent of companies worldwide plan to expand their AI investments, only 65 percent in Germany do so – a figure significantly below the global average. 52 percent of German executives feel restricted by regulatory hurdles – more than in any other country surveyed. 62 percent cited data privacy concerns as a limiting factor, and 46 percent mentioned the fear of losing control.

The economic consequences are clearly quantifiable. A study by the German Economic Institute (IW), commissioned by Google, estimated the potential of AI to increase gross value added in the manufacturing sector by up to 7.8 percent. The overall economy could grow by up to 330 billion euros through the consistent use of AI. Productivity growth in Germany had already halved before the AI era – from 1.6 percent between 1997 and 2007 to 0.8 percent between 2012 and 2019. AI would provide the urgently needed new impetus. Instead, Germany is practicing institutionalized skepticism towards technology.

The KPMG AI Index from early 2026 summed it up perfectly: The US is clearly ahead in all metrics in the global AI comparison, while Europe and Germany lag behind in rapid AI scaling despite favorable conditions. PwC found in May 2026 that only one in four German companies consistently aligns AI with growth. Strengths in governance and data are not translating into business impact. Put another way: Germany is good at writing rules. Bad at seizing opportunities.

The paradox is truly Kafkaesque: Germany is one of the few countries where a politician is punished not for bad policies, but for the alleged use of a productivity-enhancing tool, and his newspaper article is posthumously deleted. In the USA, China, Singapore, or South Korea, this would be unthinkable. Not because no one there is thinking about AI transparency, but because the fundamental societal attitude toward technology is different: How can we use it? In Germany, the dominant question is: How can we control it?

Our global industry and economic expertise in business development, sales and marketing

Our global industry and economic expertise in business development, sales and marketing - Image: Xpert.Digital

Industry focus areas: B2B, digitalization (from AI to XR), mechanical engineering, logistics, renewable energies and industry

More information here:

Expert Business Hub

A thematic hub offering insights and expertise:

Knowledge platform covering global and regional economies, innovation and industry-specific trends
A collection of analyses, insights, and background information from our key areas of focus
A place for expertise and information on current developments in business and technology
A hub for companies seeking information on markets, digitalization, and industry innovations

Between phrasing assistance and fake news: The right balance when it comes to AI

What transparency really means: A constructive proposal

The demand for transparency in the use of AI in political communication is legitimate. It is even correct. But transparency does not mean suspicion, and labeling does not mean stigmatization. There are countries that understand this difference and act accordingly.

Thuringia's Digital Minister Schütz himself said he would label the use of AI. That's a sensible position. However, the crucial question is how such labeling is embedded: as a mark of quality or as a stigmatizing label? If a statement like "This text was created with AI support" leads to automatic delegitimization, the labeling is not transparent, but destructive.

In its 2026 statement, AlgorithmWatch called for a mandatory transparency register for AI applications in public administration, as well as fundamental rights impact assessments for all AI applications in the public sector. These are nuanced, pragmatic demands. They distinguish between high-risk AI applications—for example, in law enforcement or immigration authorities—and supportive AI applications in everyday communication. This distinction is essential. Anyone who equates all AI applications morally is not thinking.

Katharina König-Preuss, a member of the Left Party in Thuringia, offered the most sober analysis of the entire discourse: She assumed that all politicians now use AI – for research or for revising texts. There is nothing inherently wrong with that. What matters is how AI is used and whether transparency is maintained. That is the correct hierarchy: first the objective classification, then the normative evaluation.

Discourse as a self-absorption program: The economics of outrage

It's worthwhile to understand the economics behind the outrage discourse. Why does an AI detector result spark a national debate? Because outrage works. It generates clicks, shares, and comments. It creates visibility for those who trigger it. It requires little effort and yields high returns – at least in the short term.

The Higher Education Forum on Digitalization precisely describes the problem in its analysis of AI detectors: A false positive result significantly harms the wrongly accused party, while the accuser faces hardly any consequences. The risk is asymmetrically distributed. Those who accuse gain attention. Those who are accused lose reputation – even if the accusation proves to be unfounded.

This asymmetry is a fundamental problem of modern media discourse. And it doesn't improve when the accusations are of a technical or algorithmic nature. On the contrary, an algorithmic judgment appears more objective and beyond question than it actually is. Anyone who wants to argue against an algorithm has to explain how algorithms work—and that is practically impossible in the context of a political headline. The combination of algorithmic authority and media outrage is particularly toxic.

In the context of the Lanz debate, Focus magazine also noted that many readers increasingly perceive political talk shows as confrontational spectacles where personal clashes are more important than substantive arguments. Trust in political communication formats is waning. And yet, parts of the mainstream reproduce precisely those mechanisms that damage this trust – because they generate short-term attention.

What's at stake: The societal costs of stifling innovation

One must ask what Germany loses if the discourse unfolds as it did in the Voigt case. The answer is not trivial.

Firstly, Germany is losing the trust of those who want to use state-of-the-art tools to communicate better, more efficiently, and more inclusively. People who use AI to compensate for language deficiencies, to express complex thoughts in a structured way, to be visible in multiple languages – they are confronted with a negative suspicion that is legitimized not by quality, but by algorithms.

Secondly, Germany is losing its appeal as an innovation hub. The Global Skills Report 2025 ranked Germany 14th in AI skills, behind Switzerland, the Netherlands, and Luxembourg. A society where using AI can be frowned upon is not an attractive place for AI professionals from around the world. Technology transfer and economic development only work in a cultural climate that embraces innovation as an opportunity.

Thirdly, Germany is missing the opportunity to set an international standard for the responsible use of AI. Instead of shaping a European model that integrates transparency and productivity, Germany is projecting the image of a country that uses transparency as a tool for stigmatization. This is the antithesis of leadership.

A population study conducted by Johannes Gutenberg University Mainz demonstrated that while the German population desires binding rules for oversight and transparency in the use of AI in politics, they want rules – not a tribunal. The difference lies in the fundamental attitude: rule-based transparency fosters trust, while moral condemnation breeds distrust.

Those who raise their hands: The sociology of knowing better

It would be incomplete to analyze the discourse on AI use without considering its social actors. For it is not just any voices that set the tone in Germany when it comes to casting suspicion on new technologies. It is a specific class of commentators, journalists, and activists who arrogate to themselves a kind of informal moral oversight – not based on democratic legitimacy, but on the basis of media reach.

These actors operate according to a recognizable pattern: They pick out a technical detail suitable for moral contrast, generate outrage, and leave the consequences to the accused. They rarely concern themselves with the actual implications of the technology they are criticizing. They rarely ask what alternatives exist and what costs those alternatives entail. They rarely consider whether their own discourse might be doing more harm than good.

This is precisely the phenomenon Johannes Volkmann addressed on Markus Lanz's show: the dominance of emotions over substance. It's not the issue itself that counts, but the gesture of outrage. The goal is not to analyze the problem, but to demonstrate moral superiority. This costs nothing and generates attention – for the person expressing outrage.

The Allensbach Institute also found in surveys that around 40 percent of Germans believe they can no longer freely express their opinions for fear of negative consequences. This is not unrelated to a culture of debate in which certain positions—such as the use of modern tools without an apologetic admission—are reflexively met with suspicion.

What Germany needs now: Pragmatism instead of moralistic politics

Germany doesn't need more tribunals on AI use. What it needs is a pragmatic debate that clearly answers three questions.

The first question is: What is the difference between AI as a formulation aid and AI as a fully autonomous text generator? This is a meaningful distinction that is both technically and normatively sound. A politician who uses AI to better structure their thoughts is behaving no differently than a speechwriter who comments on their drafts. A text that is generated entirely without any content input from the client and presented as their own work is something else entirely. This distinction is completely missing from the current discourse.

The second question is: Which labeling requirements are proportionate and practical? Labeling makes sense. But it must be embedded in a context that doesn't automatically lead to delegitimization. This requires that society's general attitude towards AI use first be normalized. As long as labeling is understood as an admission of guilt, it doesn't create transparency, but rather incentives to avoid it.

The third question is: Who is authorized to judge the use of AI? A commercial AI detector with a proven error rate is not a judge. A newspaper that republishes a political opinion piece on this basis is making a far-reaching decision based on scant data. This warrants critical scrutiny – including of the FAZ.

Heine was right – and that's not a good sign

"When I think of Germany at night, I am robbed of sleep" – Heinrich Heine's famous lines from "Night Thoughts" of 1844 describe a Germany that is its own worst enemy. 182 years later, this still sounds like a relevant diagnosis.

Germany possesses the scientific foundation, industrial infrastructure, academic potential, and economic strength to play a leading role in the AI revolution. Instead, it produces discourses that criminalize the use of AI, shame politicians over algorithmic suspicions, and pretend this is a sign of particular responsibility.

It isn't. It's the opposite: it's intellectual laziness disguised as morality. It's the privilege of those who have nothing to innovate themselves to belittle the innovation of others. And it's the collective willingness to forgo 330 billion euros of economic potential so that one can feel morally superior.

The good news: This discourse is not inevitable. It is a choice – and choices can be changed. But we need voices that call this mechanism by its name, clarify its costs, and offer pragmatic alternatives. This is not a political demand. It is a fundamental intellectual requirement for a society that wants to remain relevant in the 21st century.

🎯🎯🎯 Data-driven B2B industry hub as a quasi-in-house solution

The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business - Image: Xpert.Digital

Xpert.Digital is a data-driven B2B industry hub led by Konrad Wolfenstein . The company acts as an external, quasi-in-house solution for industrial partners, closing operational gaps in marketing, content, and sales – without requiring additional resources on the client side.

More information here: