The Cloud, Platforms, and Search Engines Are Public-Interest Institutions
Today, US tech giants increasingly capture and profit from knowledge, research, and technologies produced in European universities and public institutions. The result is a deepening digital dependency: the European Union relies on foreign-owned platforms that control not only our data, but also the way we communicate, work, think and, ultimately, live our lives.
For economist Cecilia Rikap, Associate Professor in Economics and Head of Research at the Institute for Innovation and Public Purpose at University College London (IIPP-UCL), this dependency is not merely economic: it is also a democratic issue. At stake, she argues, is who has the power to design, govern, and control the technologies.
In this interview, Rikap explains how the EU’s market-led approach has failed to secure real digital independence, why public digital infrastructure — from data centres to platforms and models — is crucial to democratic innovation, and how Europe can still regain control over its digital future.
Kirsten Van Gestel & Klara Ledroit: What is digital sovereignty?
Cecilia Rikap: Digital sovereignty is fundamentally about governance and decision-making. It concerns who gets to decide which digital technologies are developed, how they are designed, who can access them and on what terms. In addition, since digital technologies are inseparable from data, digital sovereignty is also about data governance: which data are collected, who has access to them, and which kinds of sensitive or personal data should never be extracted in the first place.
When I refer to digital sovereignty, I therefore mean democracy. I mean giving workers, citizens and public institutions the power to shape both our digital present and our digital future, instead of leaving such decisions to a handful of corporations. Contemporary digital technologies deeply affect countless actors — from workers, trade unions and civil society organisations to many others — yet they have no seat at the negotiating table.
Big Tech has acquired extraordinary power over the entire economy. We cannot afford to remain captive to Big Tech networks, argues Cecilia Rikap. Europe needs an alternative digital ecosystem.
Sovereignty classically refers to a territory. This is an exceedingly narrow definition, especially in a digitalised world. Countries in the Global South may formally exercise sovereignty over their territories, yet in practice they remain constrained by the conditions established by international lenders, since this borrowed money is essential to the continued functioning of their economies. In addition, multinational corporations routinely extract natural resources and pressure governments into adopting favourable regulations. This also undermines a territory’s sovereignty.
Associating sovereignty with a territory has become less accurate in a hyperconnected digital environment. Achieving European digital sovereignty does not mean that every component of digital technology must be produced on European soil. What matters is that the technologies we adopt are those we collectively consider as the best for us. This requires the capacity to scrutinise these technologies rigorously and develop a meaningful understanding of their inner workings.
The central issue is not whether we should produce everything ourselves, but rather that we can decide what will be produced and how. Large corporations do not dominate because they manufacture everything in-house. Instead, they outsource manufacturing while retaining decision-making power. True sovereignty implies democratising the decision-making process in the digital space.
Many people associate the technology industry with Silicon Valley and a culture dominated by profit-driven “tech bros.” Could you explain the broader issues and stakes involved?
Let’s begin with the highly sensitive issue of Internet and social media addiction. Most social media platforms originate in the United States and, to a lesser extent, in China. These platforms operate through opaque algorithms built in part on knowledge produced in Europe. Yet, Europe is denied access to these algorithms and cannot audit them properly. Ex-post content moderation alone is insufficient. By relying predominantly on content moderation, we have effectively allowed these companies to regulate themselves — much as a government, instead of maintaining a police force, would ask criminals to oversee their own misconduct. This is clearly unacceptable. Private companies cannot also be entrusted with the moderation of social media itself.
Social networks operate continuously and have become a foundational infrastructure of contemporary communication. However, we have remarkably little understanding of how they actually function at the level of their underlying code. We have no access to the key variables that determine which news we receive or not, or who can see what we publish and who cannot. We are told that this system is far more democratic than the previous media landscape, in which information was controlled by a small number of concentrated media outlets and ordinary people had no voice. What we are witnessing though is that you may post online without even being heard or read unless these companies consider it economically or politically profitable for them to spread your message. As a result, far-right messages are shared and promoted more widely than other types of content.
Technologies developed in European universities and public institutions are increasingly captured and monetised by US tech giants.
People working with these companies whom I interviewed explicitly acknowledged the existence of algorithms designed to foster addictive patterns of social media use. We also discovered that Meta and YouTube (owned by Google) had entered into agreements to maximise the amount of time teenagers spend on their platforms. Future generations — instead of being inspired by critical thought, of engaging in discussions with classmates or family members, of reading books and of developing into individuals capable of independent thought — are growing up heavily addicted to the Internet and spending their days browsing online without ever truly interacting with others or developing expected thinking skills.
With generative AI, we are now entering yet another phase: as individuals, we hand over to AI models the questions we face, thereby externalising our thinking abilities. Yet if we stop exercising our intellectual capacities, we may eventually find ourselves unable to think, respond, and act for ourselves whenever ChatGPT is unavailable. Accuracy is another major concern. Do such tools consistently provide correct answers? Certainly not. Yet people frequently overlook this and consider that the answers they receive are true and trustworthy.
Generative AI is putting our capacity to think — especially to think critically — at risk. Although AI may appear to save time and offer significant benefits, the challenges it raises extend well beyond questions of digital literacy: we need to understand both what AI can do and what it should not be used for. AI will remain largely a black box[1] for most people, and, as long as it remains so, we will be unable to understand how it works or to demystify it. I strongly believe that we need a better understanding of how corporate power shapes and influences individuals. We need to recognise the extent to which corporations have contributed to the unequal society in which we live. We also need to understand the ways in which we, as individuals, are being controlled.
A few years ago, a New York Times journalist published an account of her attempt to live without relying on Google, Microsoft, and Amazon — without any Big Tech corporation — and how she ultimately failed. Deleting a few apps may seem straightforward, but we often fail to realise that Big Tech provides many of the invisible infrastructural layers of the digital ecosystem. As a result, these companies are able to monitor much of what happens not only within the tech sphere, but also across society and the economy as a whole, occupying a position from where they can shape contemporary global capitalism, which is why this should concern us all. We need to become actively involved in the ways information is distributed, produced, and accessed.
The political left is not always seen as particularly engaged with the latest technological developments. Why should technology matter more to progressive movements?
Digital technologies are now deeply embedded in the major crises that are of concern to progressive movements. They shape migrations, labour, environmental degradation. Border surveillance systems are supplied by technology corporations, which rely themselves on digital partners such as Palantir to determine who is allowed to enter a country. From an ecological perspective, digital technologies come with a material dimension leading to the exacerbation of environmental crises, from the extraction of lithium and rare earth minerals to the massive consumption of energy and water by data centres.
Because digital technologies are inseparable from data, digital sovereignty also requires data governance.
It has also become impossible to discuss working conditions without examining the use of artificial intelligence and other forms of algorithmic management in the workplace. We need to consider the ways in which generative AI is replacing creative tasks and professions such as my own. This does not mean that all researchers, journalists, or designers are doomed. However, creative and cognitive labour is becoming increasingly automated or fragmented, leading to a segmentation of the labour market in which a minority enjoys stable and well-paid employment while a growing majority of workers are relegated to precarious or informal forms of labour. This segmentation directly contributes to the rising inequalities we are witnessing today. Fiscal systems, meanwhile, remain poorly equipped to deal with digital capitalism. The largest technology companies have become extraordinarily skilled at avoiding taxes.
Most digital services nowadays originate in the United States. They are provided by companies such as Google and OpenAI, alongside a more limited number of Chinese firms. Why is it problematic to rely so heavily on one single country — particularly the United States — for digital technologies and services?
One obvious issue relates to sovereignty and extraterritorial jurisdiction. This matter was recently highlighted during testimony by a Microsoft representative before the French National Assembly: Under the US CLOUD Act, American authorities can compel US-based companies to grant access to data, even when those data are stored outside the United States. In practical terms, it means that data hosted in European data centres can still be handed over to the US administration upon request. This is a major concern regarding the protection of European personal data, public data, and professional data. The fact that such a mechanism even exists should already alarm the European Union.
However, the issue extends far beyond questions of data access alone. Digital technologies actively shape the information we receive and the way in which we receive it. Public authorities and citizens around the world make decisions based on information. In short, we are outsourcing crucial decision-making — about how we make decisions and how we govern ourselves and our countries — to a handful of corporations located in a country entirely outside the European Union’s jurisdiction. A further aspect is the economic dimension. Dependency also creates trade imbalances. Although public debate tends to focus heavily on tariffs and the European Union’s trade surplus in goods with the United States, the US simultaneously maintains a substantial surplus in digital services. And yet this issue remains marginal in trade discussions. Moreover, the US government can require digital service providers to terminate contracts with European organisations. Such measures could completely halt the functioning of targeted entities, given the extent to which organisations rely on digital services and information systems under contemporary capitalism. This mechanism was notably employed by Washington against the International Criminal Court (ICC) during its investigations into the genocide in Gaza. At one point, the ICC Prosecutor reportedly even lost access to his email account after Microsoft cancelled his Outlook subscription at the request of the US government.
There is an obvious problem in relying on providers located in a country that claims the right to rule the rest of the world. The United States expects its allies to adopt US technologies, and those who refuse to comply are treated like foes. But turning to the Chinese digital industry would not really make a difference: Chinese Big Tech maintains close ties with the Chinese administration in ways that are similar to the relationships between the Trump administration and US tech giants. More than ever before, this situation should raise concern within the European Union.
The EU has attempted to respond to US monopolistic ambitions in the digital sphere through laws such as the Artificial Intelligence Act, the EU Chips Act, now the proposed EU Space Act… all of which are related to digital policy to some degree. Yet the EU still lags behind, primarily relying on regulation to address the problem. Why is this approach insufficient? To what extent is it flawed?
The central issue is that the EU, like many governments in the peripheries, continues to think in silos. They only see separate economies, with distinct industries and markets in each of them. And they view the world through the lens of their own economy, which they perceive in turn as confined within national borders. This mindset helps explain why the EU keeps producing so many different pieces of legislation, failing to act on the fact that we live in a highly interconnected world. This is particularly evident in the field of technology. Digital technologies are developed through vast global networks. Focusing on one single market is pointless, while it prevents policymakers from realizing what truly matters.
Strikingly, Europe is the only major region in the world that has no economic planning. In the United States, the federal government has formed strategic alliances with US tech giants. This is economic planning in practice. Together, they plan not only the development of technologies, but also their global deployment and adoption. The US administration, meanwhile, decided to acquire a 10% stake in Intel because it could not afford to see that company collapse, given its strategic position in the value chain. This is economic planning in action. It is also widely known that the Chinese government engages in economic planning as well. Digital technologies are deployed through planning, not through the dynamics of markets, and even less through the creation of competitive markets.
The EU has recently announced its ambition to position itself as the “AI continent”. Yet, as we have seen, the dominant players and monopolies remain concentrated in the United States and China. Why, in your view, is Europe so absent from the development of AI?
Let me first point out that Europe is actively involved in AI development. The fact that Europe doesn’t reap the benefits of it and appears less competitive in the field of AI does not mean that it fails to produce the fundamental inputs necessary for AI development.
The perceived competitiveness gap between Europe and the United States stems, at least in part, from the intellectual rents appropriated by US tech monopolies. These companies derive substantial profit from knowledge that has often been produced through cooperation with European universities, public institutions, and researchers. Such publicly funded knowledge isn’t simply “transferred” to large technology companies in some neutral way; rather, it is actively captured through control over digital platforms and through strategic dominance in digital value chains — a point to which I will return later.
Consequently, major technology corporations and their satellite companies sell AI-driven digital services back to Europe, even though those services rely heavily on European research and technological expertise. In doing so, they extract and relocate a significant share of the value created in Europe to the United States. Europe therefore appears less competitive — not because it lacks knowledge or technological capacity, but because it does not control how that knowledge is commercialised and governed.
Rather than building an alternative digital ecosystem of its own, Europe appears increasingly tempted to settle for the mass adoption of AI. The promise that AI will fuel economic growth — a claim that is at best exaggerated and at worst simply false — is proving extremely attractive to European governments. “AI will drive future economic gains and economic expansion”, repeat major tech companies and their allies. Yet this assumption has been contested by public figures who are by no means associated with the political left, such as Daron Acemoglu[2]. According to Acemoglu, AI-related productivity gains over the coming decade are unlikely to exceed 1%, even under the most optimistic scenario. This suggests that the large-scale adoption of AI may ultimately generate only modest increases in productivity.
With digital sovereignty, workers, citizens, and public institutions can regain the power to collectively shape the digital present and future.
I believe that the EU should place greater emphasis on solving problems. Problem-solving requires complex and targeted solutions alongside bold, interconnected policies. Such an approach demands democratic economic planning and a move away from the search for a perfect fix that could make these problems disappear overnight. Rather than focusing on a single metric — economic growth — policymakers should focus on the actual nature of the problems they face. Doing so would make it clear that alternative forms of resolution are possible. These may involve forms of degrowth or other measures that lie outside traditional economic frameworks. Economic growth is only one very specific and limited indicator. It reflects not the value that is created, but the value that is captured. Greater attention therefore should be paid to what kinds of value we should collectively want to create — and enjoy.
The EU is pouring money into the technological sector in the hope that something useful will eventually come out of it. In your paper “Reclaiming Digital Sovereignty”[3], you argue for the importance of public infrastructure. Why does public control matter so much — and why should Europe stop relying primarily on private-sector development?
In order to become truly independent from Big Tech, we need an alternative digital ecosystem. But this ecosystem cannot simply replicate the current one while just removing the large tech companies, since any company replacing them would likely be just as predatory. What is needed instead is the recognition of digital technologies as public services. This requires public data centres, public submarine cables, and public satellites, all interconnected with one another. It would also require a common platform through which digital technologies could be developed, exchanged, shared, and used.
That’s what the cloud has become today. It’s also a marketplace for digital technologies where everyone can buy technologies as services, use them — sometimes as opaque black boxes — and build new services upon them. These newly developed services are then fed back into the cloud and sold all over again.
For example, Amazon Web Services operates a marketplace quite similar to Amazon.com, except that, instead of books and consumer goods, it rents software, models, infrastructure, and many other digital services online. These services can be purchased as distinct components that are then assembled in multiple ways, much like Lego bricks, to develop information systems and applications for organisations of every kind. They are used by schools, hospitals, small and large companies, and public administrations.
The cloud is like an industrial park, but for digital services. As the development of digital solutions increasingly takes place within the cloud platforms operated by Google, Microsoft, and Amazon, these companies have acquired extraordinary power over the broader economy. Today, virtually every sector depends on digital technologies. This is why the cloud market should operate as an institution that is both public and democratic.
Europe appears less competitive — not because it lacks knowledge or technological capacity, but because it does not control how that knowledge is commercialised and governed.
This doesn’t mean that companies would have no seat at the table. We obviously live in a capitalist society in which companies employ workers and pay taxes (well, not always, admittedly, but some do). The real question therefore is what role they should play in the system.
The stack metaphor is useful in this context. A digital product is not a single, self-contained object, but a layered set of interconnected components forming a stack. Certain actors within this ecosystem therefore act as orchestrators by coordinating, connecting, and enabling interactions among the different layers — much like a shopping mall in which each store depends on the others to function effectively. Within such a model, the public sector should act as the orchestrator by providing the core infrastructure — the equivalent of the shopping mall itself — and by establishing the rules under which the various “shops” (applications, specialised software services, etc.) operate. Private firms could still maintain their own “store” within this virtual digital mall.
For instance, application layers, specialised technological services, or even parts of the infrastructure such as IoT[4] sensors could be supplied by private firms. Regarding components that should be kept under public control, we need to examine the institutional arrangements capable of making them genuinely democratic. By keeping the key elements of the system in public hands, such a model would not only ensure fair outcomes for the public, but also create fair conditions for the companies involved, since they would be able to compete on equal terms.
At the same time, some digital services operate as natural monopolies. Indeed, in some cases, having a single provider is more efficient than multiple ones. Search engines are a good example. Imagine having to run the same search query across twenty different search engines and then compare all the results — the process would become extremely time-consuming. It is therefore more practical to rely on just one (Think of AI’s evolution, where algorithms get better over time by processing more data every day). By concentrating our searches in one place, we contribute to improving the search engine itself. This network effect explains why Google’s algorithm has become so powerful: the sheer volume of data continuously enhances it. There is therefore a strong case for establishing a public search engine as a public utility, for the same reason that public services were created in the past. The same logic would also support the creation of a public marketplace for online commerce.
In Uruguay, the public telecommunications company ANTEL hosts two of the country’s three data centres and is attempting to establish itself as a provider of digital services for the public sector. In Brazil, under pressure from corporate actors, the government is negotiating with major technology firms such as Oracle to install data centres whose operation and access would be restricted to public-sector personnel. Such isolated initiatives would likely prove more effective if they were incorporated into a broader shared strategy designed to expand public control over digital technologies, at least in relation to public-sector operations.
One important point I want to emphasise is that, when I speak about public platforms and publicly governed platforms, I don’t have Europe alone in mind. To me, “public” also means “international”. Such infrastructures should be conceived at an international level, involving non-aligned governments that genuinely seek to strengthen popular democracy.
It is often argued that competitive markets drive innovation, and large technology corporations are frequently portrayed as the main engines of technological progress. Could a shift towards greater public control undermine innovation, or is this narrative misleading?
Innovation is not an individual achievement, but a collective process. Knowledge is produced by thousands of actors: universities, public research institutions, start-ups, engineers, and workers from around the world. What explains the reputation of excellence of large technological companies is not innovation itself, but rather their ability to capture, recombine, and monetise knowledge that has been produced collectively. Silicon Valley is not some magical place where innovation occurs spontaneously.
In my research, I show how companies such as Google have positioned themselves at the centre of AI research. This central position enables them to influence research agendas according to their own priorities. These corporations not only appropriate publicly funded research, but, more importantly, also shape future research trajectories in ways that consolidate their profits and intellectual monopolies. This process is far from neutral. It fundamentally transforms both the kinds of technologies being developed and the objectives that are pursued.
Introducing public control changes the logic of innovation without reducing the capacity to innovate. By creating fairer rules, organisations that were previously innovating without benefiting from their innovations would be able to continue doing so while gaining from the value they create. This applies not only to the private sector, but also to universities and public research institutions that innovate yet not for profit — for instance to treat and cure diseases, improve people’s lives, and deepen our understanding of the world in ways that assist and guide policymakers. Far from disappearing, all this creative thinking could actually be significantly strengthened once we collectively define what goals we want to pursue and determine how to achieve them.
Big Tech can steer global capitalism, which is why this should concern us all.
To make this possible, we need democratic public spaces in which knowledge can be collectively discussed and produced. We cannot leave it to the market — and to concentrated capital — to decide whether the ecological crisis should be treated as a priority. That is exactly what has happened so far, and the problem has only worsened.
We also need to break with the idea that innovation is automatically good. If we do not stop and ask ourselves what kinds of innovation we need and can realistically afford, we risk exceeding planetary boundaries and creating things that do not contribute to a good or better life for all. Consider the use of AI for warfare and surveillance. These are not desirable forms of innovation.
In a publicly governed system, innovation would still be present. In the ecosystemic alternative that my co-authors and I propose, we call for the creation of an interdisciplinary research agency governed by workers and incorporating representation of trade unions and civil society. This is essential if research agendas are to be aligned with the most pressing and sensitive issues affecting broad majorities of the population. Yet even with an agency like this, we must stop obsessing over innovation regardless of its outcomes.
To conclude, some may be under the impression that we are at the mercy of Big Tech, and yet, as you have explained, alternatives do exist. If you had to begin implementing an alternative digital ecosystem today, where would you start?
I would design an alternative ecosystem for public education. I would begin with public education because I believe we cannot afford to raise future generations entirely captive to Big Tech networks. I am referring not only to algorithms designed to generate addiction, but also to those that discourage independent thinking. It is both urgent and imperative to develop public technologies for education in an ecosystemic way. We must do so before Google or any other large tech corporation effectively takes control of children from their earliest years.
If we lose this battle, it may become impossible to teach people how to live without these companies. We should begin by raising generations that are not dependent on technologies controlled by Big Tech. It is far less costly to start with such a small public ed-tech ecosystem than to attempt to transform the entire digital technology ecosystem at once. It could then serve as a model for other sectors and certain areas of public administration. It would represent a first step towards an increasingly significant decoupling from Big Tech.
Notes:
[1] A system whose internal workings are opaque, making it difficult or even impossible to understand its functioning.
[2] Daron Acemoglu is a Turkish-US economist who, with Simon Johnson and James A. Robinson, won the Nobel Prize in Economics in 2024.
[3] Rikap C., Durand, C., Paraná, E., Gerbaudo, P. and Marx P. (2024). Reclaiming digital sovereignty: A roadmap to build a digital stack for people and the planet. Available at: https://www.ucl.ac.uk/bartlett/publications/2024/dec/reclaiming-digital-sovereignty
[4] The Internet of things (IoT) refers to devices with embedded sensors, processing ability, software and other technologies that allows to connect and exchange data with other devices and systems over the Internet or other communications networks. These devices are also known as “smart”. https://ec.europa.eu/eurostat/statistics-explained/index.php?
This article was originally published in French on the website of Lava Revue, a Belgian Marxist political magazine. English translation by Karine Gantin.
Cecilia Rikap is Associate Professor in Economics and Head of Research at the Institute for Innovation and Public Purpose at University College London (IIPP-UCL). She specialises in issues related to digital governance.
Cecilia Rikap was interviewed by Klara Ledroit, who works at the European Parliament for the Parti du Travail de Belgique, with a particular focus on industrial, energy, and environmental policies, and by Kirsten Van Gestel, who works as a policy adviser on industry and digital technologies at the European Parliament and is also active in European-level labour relations.
transform! europe is a network of 38 European organisations from 22 countries, active in the field of political education and critical scientific analysis, and is the recognised political foundation corresponding to the Party of the European Left (EL).
This cooperative project of independent non-profit organisations, institutes, foundations, and individuals intends to use its work in contributing to peaceful relations among peoples and a transformation of the present world.