Digital power and the future of democratic societies

The internet has always been governed by power hierarchies

When the internet first made its debut, hopes were running high: it would provide unfettered access to knowledge for everyone and enable people to communicate freely without relying on traditional institutions. For many, the digital space was the perfect antithesis to existing power structures.

Today we know that things are more complex. A small number of platforms now control large parts of the digital public sphere. Search engines decide which information people are made aware of, social networks steer political debates, and algorithmic systems increasingly sway decisions that have an immediate impact on people. “The internet is not a separate space from the real world,” the researcher of the Austrian Academy of Sciences emphasizes. “It depicts the same power relations that we observe in the capitalist societies.”

Digital spaces do not develop outside of societal structures. They are just as much governed by economic interests, ownership structures and political clout. Large tech companies have massive purchase these days as they provide the central infrastructures through which communication, information and economic activities are organized. The decisive question is thus not just which technologies are developed. What’s important is who decides on the rules they are subject to – and thus the impact on society these rules have.

Who decides what becomes visible?

Search engines are a case in point. For the longest time, people viewed them as technical tools: systems meant to organize the constantly growing mass of information available online. But search engines are not neutral channels. They decide which content gets visibility and which information is pushed to the background. In the digital space, visibility equals attention – and attention decides which kinds of knowledge are considered socially relevant.

Mager’s research focuses on this link between technology and society. She doesn’t just scrutinize the companies behind the search engines but the entire system shaped by economic interests, user behavior and social habits – a system that serves to stabilize the prevalent digital structures.

The example of Google shows that digital dominance is not simply the result of a successful business model. Users, companies and the media also contribute to this dominance by adapting their work processes and decisions to the logic of big platforms. “With every Google search, we further reinforce this business model.” This criticism is directed less against individual companies and more against a structure in which only a handful of privately owned infrastructures assume central functions for our information and communication needs.

Open infrastructure as a way out

Against the backdrop of the social relevance of digital information systems, we have to ask ourselves if we really want to rely on private companies to provide fundamental infrastructure. For Mager, open digital infrastructures that are not controlled by individual companies offer an alternative. A publicly available web index could, for instance, serve as the foundation for different search systems. Instead of a single company deciding which information is retrievable and how it is weighted, different providers could develop their own search tools using a shared open infrastructure. “If we had a public index, i.e., an open infrastructure, we could base a variety of different search engines on it.”

Such a structure could also create new opportunities for specialized searches. Not every inquiry is based on the same interests: scientific research, regional information, journalistic research or everyday information require different things from digital systems. “Come to think of it, it’s absurd that we are using a single search tool for online shopping, academic texts and investigative research,” the researcher points out.

Especially Europe would benefit from such alternative structures. The debate surrounding digital sovereignty is often linked to regulation, i.e., rules for big platforms and user protection. But Mager takes it a step further: it is not just about controlling existing systems but about building our own technological foundations. Digital independence is not created through policymaking alone. It also needs infrastructures on which different models can run.

When algorithms rate people

The impact of digital systems becomes especially apparent in use cases in which algorithms not only sort information but directly influence decisions affecting people. In administration, job placement or social politics, statistical models are now commonly used to process large data volumes and prepare the ground for decision-making. Such systems can accelerate processes and deliver seemingly objective assessments. But that is exactly where the problem lies hidden: algorithms handle the mathematical calculations, but they do so in a way that reflects the patterns and inequalities inherent in the data they are trained on.

Social fraud scandal in the Netherlands

A notorious example is the Dutch social fraud scandal, also known under the moniker of “child welfare fraud scandal.” The Dutch government used an algorithm to identify potential cases of fraud, resulting in the wrongful incrimination of countless families, who then had to deal with existence-threatening consequences. Especially socially vulnerable people were among those hit hardest.

This case highlights a pivotal finding of technology assessment research: it is not a system’s technology alone that decides its impact. The institutional environment, goals pursued and responsibilities assigned all make a crucial difference. “It is not the individual who should carry all responsibility but institutions that must be held accountable.”

Artificial intelligence is changing our approach to knowledge

Current discussions about artificial intelligence often follow a similar pattern: new technologies are mostly assessed according to the efficiency, process acceleration and cost-saving they enable. Mager is wary of this perspective. “The use of AI is mostly discussed within an efficiency and cost-saving paradigm.” The real question, however, is a different one: which tasks should be left to technical systems to start with? And which skills should humans continue to develop?

Especially educational institutions are currently facing these problems. Schools and universities often react to AI systems only after they have become a fixture of students’ everyday lives. The real discussion, however, should start long before that and go beyond the question of how AI can be used, and it should also address the question if it even makes sense to use it in certain situations. “Who gets to decide: do we even want AI?”

Opening up space for reflection

None of this is about a knee-jerk rejection of technology. What’s important is to create space for reflection. Teachers and students should be able to discuss the possibilities that AI brings to the table – but also its limitations. Because artificial intelligence is not only changing work processes. It is transforming our entire approach to knowledge.

When systems provide answers without allowing users to retrace the original sources, a fundamental academic and social practice is discarded: the evaluation of information. “Sources are pushed to the background more and more,” Mager points out.

This development is especially relevant in the spheres of education and democracy. Knowledge is not just about ready-made answers, it is also about the ability to categorize information, evaluate sources and compare different perspectives.

The digital future is yet to be created

Despite the shortcomings Mager has identified, she sees no reason to despair. Digital developments are not an inevitable process that societies have no agency over. Technologies do not come about in a political, economic or social vacuum. This is also the central message Mager would like to get across with her research: we can actively shape digital systems.

Open infrastructures, democratic participation and independent control can contribute to a course correction when it comes to digital technologies. The question of whether the digital transformation is happening has long been decided. Now it is crucial to determine the conditions under which it is taking place, and the social goals it should serve. “If we keep telling ourselves that the ship has sailed, we will only feel paralyzed,” Mager says.

For Europe, this means that two tasks have to be pursued simultaneously. On the one hand, existing rules for large digital platforms must be rigorously enforced. On the other hand, we need to develop our own technological solutions to counteract our growing digital dependencies. “There is no question that we need to keep pursuing regulation and legal enforcement, but at the same time, we have to keep trying to create bottom-up alternative infrastructures and niches.”

Such niches can be created in various contexts, such as research facilities, public administration, schools or open communities developing new forms of digital cooperation. Not every digital infrastructure has to follow the logic of just a few large platform companies.

The future of the internet and of artificial intelligence will not be decided in the C-level suites of tech giants alone. It will also be created everywhere that people use, question and reshape digital systems.

And this is true not only for questions of technology but also ones that concern society at large: Which kinds of knowledge do we want to promote? How much responsibility should we transfer to automated systems? Which roles should humans assume in an increasingly digitized world? “It really starts with the images in our minds. Depending on how we think about the future, we will act in the present.”

The question is thus not if digitalization will change our society. This change has long begun. The question is what kind of society we want to create with the help of digital technologies. The digital future is not a technological inevitability. It is a task to be addressed by policymakers and democratic societies.

About the researcher

Astrid Mager is a science and technology studies scholar specializing in the interconnection of the internet, technology and society. She studied sociology and communication sciences at the University of Vienna, where she completed her doctoral thesis on the internet as a source of health information in 2010.

Since 2012, she has been working at the Institute for Technology Assessment (ITA) at the Austrian Academy of Sciences (ÖAW), where she researches topics like search engines, algorithms, digitization, democracy and automated decision-making systems. Her habilitation project titled “Algorithmic Imaginaries. Visions and values in the shaping of search engines” (2016–2022) was funded by the Austrian Science Fund (FWF) in the framework of the Elise Richter fellowship. From 2018 to 2026, she was a member of the Austrian Academy of Sciences’ Young Academy, and she currently leads the ÖAW Commission “Democracy in digital societies” (DEMGES) together with Barbara Prainsack.