“Democracy needs many voices” is a week of action organised by the Federal Association of Digital Publishers and Newspaper Publishers, of which Klambt Verlag is a member. Delius Klasing, as part of Klambt Verlag, is also taking part in this broad alliance and is publishing selected articles as part of this initiative.
Mr Andree, you have once again explored the digital world for your new book. What alarmed you the most in the process?
The fact that this massive monopolisation is, mathematically speaking, virtually identical to what it was six years ago – even though the EU’s Digital Markets Act was actually intended to regulate precisely that. That didn’t surprise me, though. After all, we see every day how independent providers are being increasingly squeezed out by the tech giants.
Yet the choice on the internet seems limitless. Millions of websites, apps and media services are just a click away. Just how diverse is our digital world, really?
That is precisely why the problem is so difficult to spot: with 18 million different services on offer, diversity seems more than assured. However, what matters is not what theoretically exists somewhere, but what is actually being used. And here we find that it is primarily the tech monopolies that are being used extensively.
We measure this concentration using the Gini coefficient. It ranges from 0 to 1: zero indicates perfect equality, whilst one indicates maximum concentration. When it comes to the distribution of attention in digital media, we arrive at a value of 0.987. This indicates an almost unimaginable degree of monopolisation.
Google, YouTube, Instagram and other platforms are so big partly because millions of people use them of their own free will and clearly find them useful. Why does this commercial success become a problem for democracy?
Media monopolies must be rejected on principle. This applies where they are under state or autocratic control. However, it also applies where private corporations control the monopolies.
Imagine if a mega-entrepreneur had controlled all editorial media in Germany during the 20th century. In response to criticism of this monopoly, he might have said: “I have thousands of journalists of various political persuasions who work completely independently. Editorial and management are separate – what’s the problem?” We would quite rightly never have accepted that. And we should be just as unwilling to accept it in the digital sphere. In principle, it makes no difference whether oligarchs control public opinion through direct influence or through algorithms. Media monopolies have no place in democracies.
But anyone can leave Google or Instagram, use other services or access information directly from various sources. Aren’t you overestimating the power of these platforms?
Digital media account for around 60 per cent of mass media attention and thus form the foundation of political opinion-forming in our democracy. The platforms play a key role in determining what people are shown – primarily through algorithmic feeds and now even through AI-generated summaries.
So, if people are, as it were, having content from monopolists fed to them directly from above, how are they ever supposed to know what they might be looking for? That is why the call in media policy for editorial content to be ‘findable’ falls far short of the mark.
The public debate often centres on disinformation, hate speech and illegal content. However, you see the fundamental problem as lying in the platforms’ business model. Why?
These two aspects are interlinked. For me, the core problem is the liability exemption: we have allowed economic actors to emerge who make immense profits from business models for which they are largely exempt from liability.
To this day, platforms can still make money from content that is a criminal offence, such as incitement to hatred, defamation, racism or Holocaust denial. As far back as the 1990s and 2000s, criticism of this was dismissed with the argument that people were trying to ‘censor’ the internet and restrict freedom of expression. I believe that is wrong.
Freedom of expression can also be ensured whilst maintaining anonymity, without allowing criminal offences to be monetised. We have made more progress in other areas: on the roads, we drive without using our real names and enjoy complete freedom of movement – but anyone who breaks the law must take responsibility for their actions.
Now artificial intelligence has come into the picture. Google and other providers are increasingly answering questions themselves, rather than directing users to the original sources. What difference does this make?
Our latest measurements paint a dramatic picture. Google, for example, is able to massively increase session duration. This is undoubtedly linked to ‘zero-click content’: initially it consisted of overviews and snippets, but now also includes AI-generated summaries. In the process, the editorial providers of the content are effectively being dispossessed – whilst Google grows even fatter at their expense.
What does that mean in practical terms for journalism?
We are currently experiencing a massive crisis of press freedom. The *Spiegel* affair of 1962 pales into insignificance by comparison. Press freedom is not only under threat when state actors intervene, as they did back then. It is also at risk when journalism, under the conditions of digital monopolies, loses its visibility and thus its voice – and, under the conditions of polarising algorithms, also becomes restricted in terms of content. Unlike in 1962, however, there is no protest whatsoever. Nowhere.
External circumstances are one thing. But what can newspapers, broadcasters and other journalistic media do better themselves? Have the media also made mistakes in their dealings with these platforms?
Under monopoly conditions, the problem cannot be solved through business initiatives or innovation. It is astonishing that the editorial media themselves hardly report on the destruction of press freedom. Journalists often say that one must not ‘side with a cause’ – but I consider it fundamentally wrong not to ‘side with’ press freedom. That is precisely what happened in the *Spiegel* affair – fortunately for Rudolf Augstein. Finally, editorial media could make much greater use of alternative platforms such as Mastodon or BlueSky – and actively refer to them in their articles.
With the Digital Markets Act and the Digital Services Act, Europe has introduced regulations for large platforms. You consider this to be insufficient. What specific action should policymakers take?
The tragic thing is that, in principle, monopolies could easily be opened up to competition through regulation. We have demonstrated this ourselves in Germany with Deutsche Telekom. Similar mechanisms could be implemented, for example, through the State Media Treaty.
Together with Prof. Nikolaus Peifer, I published a paper on this subject in 2024, in which we set out how this could be implemented in practice. However, those responsible for media policy lack the courage to tackle this root cause of tech monopolies.
Finally, let’s look to the future: when we look back in 2035 – how would you know that we have regained the digital public sphere?
We will be able to tell whether Germany is still a democratic state by this. For, in the long term, democratic sovereignty is only possible with digital sovereignty in the field of the media. ‘All state authority derives from the people’, as the Basic Law states. But the people obtain their information from the media. Digital media are dominated from outside, from Silicon Valley. This means that the democratic community has long since lost its own voice in shaping its democratic forum. Yet in a public sphere controlled by external forces, democratic sovereignty also collapses.
About the author: Martin Andree is a media studies scholar and lectures at the University of Cologne. For many years, he has been exploring how Google, Meta and other tech giants are shaping our digital public sphere and what their growing power means for the media and democracy. In his book *Krieg der Medien* (2025), he examined the consequences of platform power for media freedom and democracy. For *Measuring the Digital World* (2026), he analysed usage data to determine how our attention is actually distributed online and just how dominant Big Tech has become.
Find out more at democracy-needs-lots-of-votes.de