In January, the European Commission released the first draft of the Digital Networks Act (DNA), proposing changes that would fundamentally alter the functioning of the internet interconnection system – with profound negative consequences for free expression and access to information.
The internet’s interconnection system, the web of agreements through which different networks exchange traffic, is one of its great success stories. For decades, internet service providers (ISPs), content networks, and other operators (such as telecommunication companies) have negotiated voluntary agreements to exchange traffic directly between networks, keeping data flowing efficiently across the globe. No central authority mandates these agreements. No regulatory body approves the terms. The system works through commercial negotiation, technical coordination, and mutual benefit. It has scaled from connecting university networks to supporting billions of users and services, all with careful regulatory oversight that only steps in when needed.
The European Union’s Digital Networks Act (DNA) proposes to change this delicate balance. Not through explicit regulation of interconnection – the current draft is careful to avoid that language – but through what it calls ‘ecosystem cooperation’ and ‘conciliation mechanisms’ for traffic exchange. These provisions, buried in articles on network resilience and dispute resolution, treat interconnection as a regulatory problem requiring further oversight and mandatory mediation. On top of that, the DNA’s approach to regulatory simplification raises concerns. While streamlining procedures for satellite operators and large cross-border networks, for example, it does little for smaller stakeholders like community networks and alternative ISPs. Moreover, the rush to embrace deregulation often uses simplification as cover for centralisation – absorbing technically independent bodies into political structures, with transparency as the first casualty. Without symmetrical simplification that empowers smaller actors, the DNA risks reinforcing market concentration rather than addressing it.
‘The internet’s interconnection system works because it is flexible and decentralised. The DNA’s proposals might sound reasonable, but they misunderstand how traffic exchange actually operates, and risk replacing a proven system with bureaucratic oversight that favours incumbent players.’, Dr. Corinne Cath, Head of Digital at ARTICLE 19
The Internet Society’s detailed analysis of the draft asks a crucial question hanging over the DNA: what problem is this solving? We will use this question to discuss the various answers and why they come up short from a human rights perspective.
No market failure, no justification
European internet interconnection works. Networks connect, traffic flows, and users access content from around the world with remarkable reliability. When problems arise – network outages, routing issues, capacity constraints – operators resolve them through existing commercial and technical channels.
The Internet Society and other experts, including dr. Konstaninos Komaitis, examined the DNA’s justifications for intervention in the current market and found them wanting. The proposal gestures towards concerns about ‘fair traffic exchange’ and ‘network sustainability’ without demonstrating that current interconnection arrangements actually harm competition, innovation, and network investment. In fact, the evidence points the other way: voluntary peering, where networks exchange traffic by mutual agreement, has enabled massive traffic growth whilst keeping costs down. European internet exchange points are among the world’s most successful, handling enormous volumes of traffic whilst remaining open and competitively neutral. Content delivery networks (CDNs) have flourished, bringing content closer to users and improving performance across the continent.
The answer to why the Commission would intervene in a functioning market lies not in a technical necessity but in a long-running political campaign by large telecommunications operators. They found a particularly receptive environment under the previous Commission, where a former telecom executive, Thierry Breton, was the Commissioner for Internal Market. Under the term ‘fair share’, for over a decade, major European telcos have argued that application- and platform-layer internet services (such as streaming platforms and social media) should compensate network operators for the traffic they generate. EU consumer and telecom authorities including BEUC and BEREC have repeatedly rejected these proposals, recognising that they fundamentally misunderstand how the internet works and would harm rather than help European connectivity.
Fair share by another name
The DNA does not explicitly mandate fair share payments. It does not need to. Instead, it creates procedural mechanisms through which dominant network operators can pressure other actors into financial settlements.
The ‘conciliation’ framework allows operators to claim that traffic patterns create problems requiring regulatory mediation. This is cynical, as it reframes ordinary commercial traffic flows as a regulatory problem, despite the absence of any technical failure or market breakdown. The ‘ecosystem cooperation’ language implies that current arrangements are somehow insufficient. Together, these provisions construct interconnection as something contentious rather than cooperative and create venues for ‘manufactured’ disputes.
Consider the practical implications: a large telecommunications operator claims that a streaming service creates ‘excessive traffic’ on its network. It files for conciliation under the DNA framework, arguing that the content provider should contribute to network upgrades. Never mind that the operator is already obligated to maintain sufficient capacity for the traffic its customers request. Never mind that those customers already pay the operator for internet access specifically to reach services like streaming platforms. Never mind that the content provider already pays for its own network connections and CDN infrastructure.
Under the DNA’s conciliation mechanism, the regulator must mediate. What position does the streaming service negotiate from? If it refuses to pay, it risks degraded service quality as the operator (i.e. the telco) manages its capacity. If it pays, it sets a precedent for every other content provider to face similar demands. If it is a small or medium-sized service, for example an independent news outlet, a European video platform or a community media project, it may lack the resources to negotiate at all, leaving it at the mercy of whatever terms the operator demands.
This is not speculation. Similar dynamics emerged during previous fair share debates, where operators sought payments from services whose content users actively chose to access. The current case against Deutsche Telekom focuses on precisely this practice, where certain services experience degraded quality unless they pay for preferential treatment, as documented in the netzbremse.de campaign. Now, the DNA would give these demands procedural legitimacy through formal dispute resolution mechanisms. What operators could not achieve through transparent policy debate, they could accomplish through case-by-case regulatory processes.
From cooperation to rent-seeking
The deeper problem is what these mechanisms do to interconnection as an ecosystem. Currently, networks connect because it is mutually beneficial: both sides gain from exchanging traffic efficiently. These arrangements are voluntary and can be terminated if they stop working for either party. This flexibility has allowed the internet to evolve rapidly, adapting to new technologies, traffic patterns, and business models without regulatory permission.
Introducing the type of mandatory conciliation and regulatory oversight proposed in the DNA transforms this dynamic. Interconnection stops being about mutual benefit and becomes about leverage. Large operators gain new tools to extract payments or favourable terms. Smaller networks face uncertainty about whether their peering arrangements, under which networks interconnect to exchange traffic without central mandates, could be challenged. Content providers, big and small, must factor potential network fees into their business models. We have already seen this pattern play out in South Korea, where mandatory payment schemes have led content providers to reroute traffic to avoid fees or abandon smaller markets entirely when costs exceed demand. This fragments the internet without delivering better service, harming especially the smaller players who are most likely to innovate.
Under such regulation, freedom of expression and access to information becomes a paid-for-service. The Internet Society has warned that this shift from voluntary cooperation to regulatory rent-seeking would undermine competition, discourage smaller operators and content providers, and shift costs to users without improving network performance or user outcomes. They describe what economists call ‘regulatory capture’, when procedural mechanisms designed to address market failures instead become tools for incumbents to protect their position and extract rents.
The freedom of expression stakes
At ARTICLE 19, we recognise this pattern. When dominant actors gain procedural leverage over information infrastructure, freedom of expression suffers; not just through content censorship, but through structural barriers that determine who can afford to reach audiences. Those threats are articulated particularly strongly by an Austrian NGO Epicenter.works whose analysis highlights concrete legal risks for net neutrality in the proposed reform. They warn that the DNA’s provisions could create legal loopholes that undermine existing net neutrality protections, allowing network operators to discriminate between different types of traffic under the guise of network management and resilience.
European human rights law protects not just what people can say, but their ability to access diverse information sources. Article 10 of the European Convention on Human Rights covers the right to ‘receive and impart information and ideas without interference by public authority’. The European Court of Human Rights has made clear that this protection extends to the technical infrastructure enabling communication. Measures that create differential access, impose gatekeeping, or favour certain speakers over others can constitute Article 10 interferences even if they don’t directly target content. The DNA’s interconnection mechanisms pose exactly this risk. If streaming services face payment demands, which ones will be able to afford them? Major US platforms with massive resources certainly will. European competitors trying to build alternatives might manage if they can raise enough capital. Independent media outlets serving specific communities? They are unlikely to have the resources. Community networks providing local content almost certainly will not.
The result would be a structurally tilted playing field, where capacity to pay determines capacity to reach audiences. Large, well-resourced actors will be able navigate the system; smaller, public-interest-oriented actors will face barriers. This wouldn’t just harm competition, it would undermine the information pluralism that freedom of expression requires.
‘When economic leverage over infrastructure translates into leverage over which voices reach the public, we are no longer just talking about telecoms regulation. We are talking about who gets to participate in public discourse.’ Dr. Raquel Renno, Senior Policy Officer and connectivity expert at ARTICLE 19.
What should happen instead
The path forward is not complicated: leave interconnection alone. The system works. It has demonstrated remarkable adaptability and resilience precisely because it has not been subjected to these types of ‘manufactured’ regulatory mandates and dispute mechanisms. If the Commission wants to strengthen European connectivity, it should focus on areas with genuine market failures: extending infrastructure to underserved regions, supporting alternative operators and community networks, and ensuring open access to essential facilities.
These are real problems requiring real solutions. Creating procedural mechanisms for rent-seeking in a functioning market is not amongst them. As the Internet Society puts it, no clear market failure has been demonstrated to justify such intervention. After decades of evidence that voluntary peering supports innovation and competition, the burden of proof lies with those proposing regulatory disruption. At the moment, the DNA does not meet that burden.
For freedom of expression, the stakes are clear. An open internet does not mean an unregulated one. It does, however, mean one where regulation addresses genuine problems whilst protecting the conditions that enable the diversity of voices. The DNA’s interconnection provisions do the opposite: they manufacture problems in a functioning system whilst creating new tools for dominant actors to limit competition. That is not just bad internet policy. It is a threat to the structural conditions that make European information pluralism and diversity possible.