The “interview” format of the Digital Society Lab aims to give a voice to a diverse group of experts on digital issues, all of whom speak in their own capacity. The analyses and opinions expressed are those of the interviewees alone and should not be interpreted as an official position of the ANCT’s Digital Society Program. This interview was conducted by François Huguet on behalf of the Digital Society Lab. Huguet is a researcher in the digital humanities (PhD, Institut Polytechnique de Paris) whose work focuses on investigating, documenting, and promoting digital domains of public interest.
Foreword
Francesca Musiani is a Franco-Italian sociologist and researcher specializing in the study of digital infrastructure and Internet governance. A research director at the CNRS, she co-founded in 2019 and currently directs the Center for Internet and Society, where she conducts research at the intersection of sociology and Science and Technology Studies (STS).
Her research focuses in particular on distributed architectures (such as peer-to-peer networks), technical regulation of the Internet, governance of digital infrastructure, and the political and social issues related to decentralized computing systems. She is particularly interested in how technical choices incorporate forms of power and regulation, helping to shape digital practices and standards. She regularly contributes to interdisciplinary projects and discussions on the future of the Internet and its regulation, emphasizing a critical and sociotechnical approach to technology.
Understanding Policy in Digital Infrastructure
François Huguet: In your book *Politics in Networks: Digital Power and Infrastructure*, published last February by C&F Éditions, you show that the Internet is not merely a technical space but also a site of power and governance. How does this perspective change the way we think about public policy in the digital realm?
Francesca Musiani: Indeed, the starting point for this book—and for my work in general—is to view the Internet differently: as both an economic sector and a public space, but also as a genuine political infrastructure. I seek to highlight this value as a techno-political infrastructure: how technical choices and IT architectures—the ways in which computer systems are designed, organized, and structured, from both technical and functional perspectives—produce real effects of power. From this perspective—which could be summarized as the “pipeline perspective”—we can conceive of public policy in a different way. For example, we will not focus solely on the types of content and data circulating on the Internet or on how to regulate them, but will also concentrate on the very infrastructure that enables them to circulate.
We need to focus on the software rules that govern this system and the policy choices that are “encoded” in these infrastructures. The advent of the cloud has allowed us to pay a little more attention to these choices: Where is our data “physically” located? Under what legal frameworks is it governed? Who controls the data centers where it is stored? This inevitably changes the way we think about public policy: we no longer react after the fact; we no longer regulate only the content, but we work on the infrastructure itself that allows us to see these flows.
FH: You talk about “governance through infrastructure”—that is, the fact that technical choices can have political consequences. You cite, in particular, the Domain Name System (DNS), Bitcoin, Signal, and the Internet in Russia as examples. How can this idea help policymakers better understand today’s digital challenges?
FM: Let’s clarify here that the book you mentioned earlier does not address the issues surrounding AI regulation, which have been the subject of much discussion since last year. This work draws on research conducted over the past 15 years, prior to the advent of “mass-market” AI and the myriad issues it has raised. But to get back to the topic at hand, I believe that demonstrating that Internet governance is not limited to laws or institutions—but is also embedded in the technical infrastructure itself (protocols, standards, architectures)—allows us to understand the power dynamics of the digital world in a different light. Once we understand this, we begin to view technical choices differently, to avoid underestimating them, and to realize that IT architectures embody, by design (from the very outset), different worldviews. In doing so, we can ask different questions: Where are the technical systems we use created? Who manages them? How? Establishing control and regulation therefore plays out—equally—at the infrastructure level. This is what I’ve worked on extensively with my colleague Laura DeNardis; together, we continually demonstrate that the Internet is a space where power flows through technical systems, and where regulation is often invisible but very concrete.
In the book *Politics in Networks: Digital Power and Infrastructure*, I draw on examples that allow me to highlight the sets of constraints and opportunities embedded in IT architectures. For example, I discuss the DNS (an essential Internet service that translates domain names into IP addresses), a key element of “governance through infrastructure.” The DNS has a primary and explicit regulatory function: to serve as the Internet’s “directory”; however, its technical characteristics have led several actors to attempt to use it as a tool for censorship, for making online resources inaccessible, and for enforcing intellectual property rights. I then take the example of Bitcoin: a form of regulation without states, but not governance without power dynamics—quite the contrary: Bitcoin eliminates certain types of intermediaries while giving a much more significant role to others. I also discuss the Signal messaging protocol, which shows us that encryption—when “encapsulated” within an IT service architecture—protects communications but makes legal regulation extremely complex. Finally, I discuss the Internet in Russia based on the research we conducted as part of the ANR ResisTIC project. Russia is, in fact, a “laboratory of digital resistance” but also a place where the Internet has been completely reconfigured to be far more controllable, sovereign, and to become a direct tool of state power.
Infrastructure as a Forum for Collective Decision-Making
FH: The “multi-stakeholder” model is often presented as a political innovation in Internet governance. Do you think it is a relevant model for thinking about digital infrastructure of general interest?
FM: Yes, the multi-stakeholder model is based on bringing together a diverse range of actors—from the public, private, technical, and civil society sectors—to discuss issues around the same table. In this sense, it marks a break from a strictly state-led model. However, this form of governance is not necessarily fully democratic. Numerous studies have highlighted its limitations: there are significant power imbalances among stakeholders, with large corporations possessing substantial resources (particularly in terms of lobbying) carrying more weight than smaller actors. Furthermore, the technical nature of the issues acts as a barrier to entry that limits participation. Thus, while these mechanisms appear relevant as forums for coordination and discussion, their transformation into genuine decision-making forums remains incomplete. In short, they represent a preferable alternative to purely state-led regulation, though they do not constitute a fully developed democratic model.
FH: In your research, you emphasize sociotechnical controversies as moments when infrastructure becomes visible and a subject of discussion. Can these controversies be viewed as forms of democratic deliberation surrounding technologies?
FM: On issues such as net neutrality, encryption, and content moderation, controversies play a crucial role: they bring to light infrastructures that are usually invisible. They force us to spell out technical choices that are often implicit, to highlight different levels of engagement, and to clarify the positions of stakeholders as well as the reasons behind their decisions. In this sense, they open up a space for public debate about technology. However, these forms of deliberation have several limitations. They remain unevenly accessible, particularly due to the technical nature of the issues. While they represent moments of politicization—where issues such as net neutrality become subjects of public debate—this politicization remains partial and context-specific. Topics such as encryption or the fight against disinformation clearly illustrate these recurring controversies, where technical, political, and social issues intersect, without, however, leading to fully inclusive forms of democratic deliberation.
Digital Technologies for the Public Good and Infrastructure
FH: These days, there’s more and more talk about digital infrastructure for the public good. Based on your work on Internet infrastructure, how would you define what a digital infrastructure for the public good might look like?
FM: I believe we could propose a definition based on three main dimensions. First, the accessibility of the infrastructure, which implies non-discriminatory openness, embedded in the technical choices themselves. Second, pluralistic governance—not controlled by a single actor, but involving a diversity of stakeholders. Finally, the “by design” incorporation of certain fundamental rights, such as transparency, privacy, and freedom of expression. It is precisely to analyze these dimensions—at the intersection of the technical, political, and social spheres—that I draw on Science and Technology Studies in my work!
FH: Your work shows that certain infrastructure elements are becoming points of control over power on the Internet. How can we prevent these points of control from being co-opted by specific economic or political interests?
FM: In my work, I show that certain nodes on the Internet concentrate forms of power. To limit the risks of capture, three main strategies can be employed. First, diversifying infrastructure to prevent technical monopolies. This involves maintaining technological pluralism by ensuring the existence of alternatives and avoiding excessive dependence on a few dominant players such as Google or OpenAI. Second, transparency in rules and decisions. This involves, in particular, audit mechanisms—whether driven by citizen-led initiatives, such as France’s Observatoire des algorithmes publics (ODAP), or by more institutionalized public mechanisms. The goal is to make technical choices and their effects transparent. Finally, a third lever lies in the return of forms of public regulation, through a legal framework focused on these control points, to ensure that they remain compatible with the public interest and the law.
FH: Is it possible to envision digital infrastructure designed from the outset with values that serve the public interest—for example, in terms of transparency, the protection of rights, or universal access?
FM: Yes, it’s possible! This is often referred to as a “by design” approach. This involves, in particular, prioritizing interoperability over siloed approaches, ensuring the transparency and auditability of systems (especially algorithmic ones), and using encryption to protect communications. The challenge is to ensure that these values—universal access, protection of rights, and openness—are built directly into the technical architecture of the systems, rather than added as an afterthought.
Control Infrastructure vs. Resistance Infrastructure
FH: In your book, you show that infrastructure can be used both to control and to resist. How does this tension manifest itself today in the digital ecosystem?
FM: This tension is manifesting itself today on several levels. On the one hand, we are seeing increased state control. This is particularly evident in authoritarian regimes—such as Russia, China, or Iran—through surveillance, filtering, and blocking measures. But these dynamics are not unique to them: democratic states are also implementing forms of infrastructural control, as illustrated, for example, by the French authorities’ blocking of TikTok in New Caledonia in 2024, which I mention at the beginning of the book.
Furthermore, this tension is also reflected in the concentration of power in the hands of private actors—particularly major platforms and technology companies—which are playing an increasingly significant role in regulating online communications. In response to these control dynamics, forms of technical and political resistance are emerging. Decentralization allows for the redistribution of power by avoiding single points of control, even if it in turn complicates governance and the exercise of authority. Encryption, for its part, is a key tool for protecting communications and limiting surveillance capabilities. Thus, the contemporary digital ecosystem is marked by a constant tension between centralization and the distribution of power, control and autonomy, in which technical choices play a decisive role.
FH: Certain technical projects—such as encryption or decentralized architectures—are sometimes presented as tools for empowerment. Can we really “engage in politics” through technical architecture?
FM: Technical architecture opens up certain possibilities while limiting others. It guides—or even dictates—certain behaviors. When we make these architectures visible and clarify the choices they embody, it becomes possible to redistribute power. However, this does not mean that technology replaces politics: it coexists with it. We must therefore avoid any form of technosolutionism. Furthermore, users always reinterpret technical systems, and power dynamics do not disappear. Technology must thus be understood as a political tool, not as an autonomous solution.
Digital Futures for the Public Good
FH: Given the current trends in the Internet—platform consolidation, geopolitical issues, digital sovereignty—what do you see as the main challenges to ensuring that the Internet continues to serve the public interest? If you had to identify one priority area today for building an Internet more aligned with the public interest, what would it be?
FM: The first is the concentration of platforms, which Laura DeNardis has described as the “privatization of Internet power.” An increasing share of regulation is now carried out by private actors, which raises significant democratic concerns.
The second challenge stems from the political fragmentation of the network. The Internet is no longer merely a global space: it is being reshaped by state, economic, and technical forces, raising the risk of a “splinternet” in which increasingly compartmentalized spaces coexist.
The third challenge concerns the lack of transparency in the infrastructure. The technical workings of the network remain largely unknown, which limits the capacity for democratic oversight. Hence the importance of developing audit mechanisms capable of conducting concrete examinations of the systems and code.
Finally, a fourth challenge is that of technological dependence. Governments must find ways to maintain a degree of autonomy without losing their capacity to act or isolating themselves. In this context, the top priority is to strengthen the capacity for public action regarding digital infrastructure. This begins with training decision-makers on technical issues, which are still not sufficiently understood. It also involves documenting critical infrastructure by supporting monitoring centers capable of identifying points of concentration and vulnerabilities. Finally, it involves creating genuine decision-making forums regarding these infrastructures by better anticipating investment and planning choices, rather than intervening after the fact.
