Free speech and censorship

Free societies trust citizens to think, speak and argue for themselves. The answer to bad ideas, political disagreement and alleged misinformation should be more open debate, not a permanent architecture of censorship, pre-emptive “resilience” and institutional control over what people are allowed to see and say.

The EU is rapidly constructing an unprecedented system for governing Europe’s information environment. The Digital Services Act (DSA) gives regulators new powers over major online platforms and requires the largest platforms and search engines to address risks to electoral processes and civic discourse. The ‘European Democracy Shield’ goes further, bringing together platform regulation, fact-checkers, civil-society organisations and anti-disinformation initiatives into an increasingly integrated infrastructure for monitoring and responding to political information. In doing so, what is presented as a shield for democracy risks becoming a shield against democracy itself.

Behind this architecture lies a fundamental change in how European institutions understand the citizen. Post-war European democracy rested on the assumption that individuals are autonomous agents capable of judgement. Fundamental rights therefore primarily protected citizens against excessive state intervention. Freedom of expression protected not merely correct or respectable opinions, but also misguided, offensive and controversial ones, because democracy presupposed that citizens could encounter competing arguments and reach their own conclusions.

That assumption is increasingly being reversed. Citizens are portrayed as vulnerable to algorithms, manipulation, misinformation and disinformation. The task of government consequently expands from protecting individual freedom to protecting individuals against influences that might lead them towards supposedly harmful beliefs or decisions. Rights that once functioned primarily as shields against public power risk becoming justifications for preventive intervention.

The ‘Democracy Shield’ demonstrates where this logic can lead. Organisations presented as independent participants in civil society can simultaneously receive EU funding and acquire important functions within systems of fact-checking, risk assessment and information governance. Networks such as EDMO and EFCSN become integrated into a regulatory ecosystem alongside EU institutions and digital platforms. Citizens can vote governments out of office; they cannot vote out the NGOs, foundations and fact-checking organisations increasingly involved in determining what constitutes misinformation, manipulation or legitimate political speech.

The expansion of Foreign Information Manipulation and Interference (FIMI) makes this problem even more acute. The European External Action Service (EEAS) defines FIMI broadly as predominantly non-illegal, intentional and coordinated manipulative behaviour by powers such as Russia to influence political processes, involving state or non-state actors and their proxies inside or outside Member States. Foreign interference is a genuine problem. But the danger of FIMI arises when the language developed to identify hostile foreign operations migrates into interpreting legitimate domestic politics. Populist movements, for example, and opposition arguments that challenge EU Commission orthodoxies on the environment, immigration or energy, are presented not as legitimate expressions of internal political disagreement, but as the result of external manipulation and the amplification of disinformation. Political provenance begins to compete with political argument, shifting the discussion from whether an idea is right or wrong to who might benefit from citizens believing it. FIMI thereby risks providing a vocabulary through which political differences can be delegitimised by association rather than answered through democratic debate.

Even more revealing is the growing enthusiasm for ‘prebunking’: attempting to make populations resistant to supposedly misleading narratives before they encounter them. Instead of trusting citizens to evaluate competing arguments, institutions seek to shape how information will be interpreted in advance. The citizen increasingly becomes an object of preventive governance whose informational environment—and eventually cognitive responses—can legitimately be managed.

This creates an unavoidable problem of political power. ‘Disinformation’, ‘manipulation’, ‘harmful content’, FIMI and ‘illegitimate influence’ are not politically neutral categories that enforce themselves. Someone must decide what they mean. Who decides? In an increasingly polarised Europe, handing such decisions to unelected executive institutions, regulators, EU-funded organisations and digital platforms creates enormous potential for political disagreement to be reclassified as an information problem.

The danger is a system in which democratic uncertainty itself becomes something to be managed. Dissent can become ‘disinformation’, disagreement ‘polarisation’, populism evidence of manipulation, and unpredictable political behaviour a threat to democratic ‘resilience’. Yet democratic pluralism necessarily involves persuasion, conflict and uncertainty. Once institutions acquire responsibility for managing political debate in pursuit of safer or more resilient democratic outcomes, the distinction between protecting democracy and controlling the conditions under which democracy takes place begins to disappear.

Europe needs a different approach. Protecting children, privacy and citizens against genuine coercion or exploitation does not require abandoning confidence in human autonomy.

Freedom of speech is therefore not merely one democratic value among many. It rests on the deeper democratic assumption that citizens possess moral independence: that they are entitled to hear arguments, confront falsehoods, encounter offensive ideas, and decide for themselves what to believe. Increasingly, however, the emerging European ideal appears to be not simply freedom of speech but freedom from speech—freedom from exposure to information classified as misleading, manipulative, polarising, or harmful, justified in the language of safety and democratic resilience. That is a profound inversion. A right originally intended to protect citizens from those exercising power risks becoming a responsibility exercised by those in power to protect citizens from ideas.

FSI Europe will investigate the DSA, the ‘European Democracy Shield’, AI regulation, content moderation, fact-checking networks, FIMI, prebunking and behavioural governance, exposing where legitimate regulation becomes political or informational control. But we will also make the positive case for freedom in the digital age: technological change does not make human autonomy obsolete. It becomes even more important when the European Commission—an unelected executive institution—presents itself as the guardian of democratic resilience while helping to build an infrastructure intended to manage the information environment in which democratic choices are made. Democracy, if it is to mean anything, begins with the opposite premise. A free society depends on citizens being trusted to think for themselves; a democracy that no longer trusts its citizens will eventually stop treating them as sovereign.