From Content to Structure in Democratic Self-Defense

You-Hao Lai

You-Hao Lai, S.J.D. Candidate, The George Washington University Law School; LL.M., Harvard Law School (2023). Deputy Director, Democratic Governance Program, Research Institute for Democracy, Society, and Emerging Technology (DSET), Taiwan.

     In late May, constitutional scholars gathered at Aristotle University in Thessaloniki for an IACL roundtable on militant democracy and the rule of law. Much of the discussion tracked the field’s defining issues: anti-democratic parties, hate speech, extremist movements, and the danger that a democracy may be eroded by the very process of defending itself. Yet what struck me, as a scholar from Taiwan, was an emerging concern moving to the center of the debate: democratic self-defense against threats that originate not from within, but from abroad. This concern came into sharp focus on the conference’s second morning, when the head of Romania’s Constitutional Court recounted the court’s decision to annul the first round of the 2024 presidential election in the face of foreign-orchestrated digital manipulation.

     Taiwan knows this terrain firsthand: the island nation faces some of the heaviest foreign information manipulation in the world. But it is also a democracy forged out of nearly four decades of authoritarian rule, deeply wary of state censorship. Out of that tension, Taiwanese civil society built a widely admired defense of fact-checking and media literacy. For years, it proved effective. These bottom-up efforts, however, are increasingly outmatched by floods of algorithmically driven propaganda deployed by a state-level actor. From Romania to Taiwan, frontline democracies urgently need a theory of how to defend against a new kind of threat amplified by digital technology—a defense that neither waits for a court to act after the fact nor rests on civil society alone.

Two Blind Spots

     But the existing theory is not yet equal to the task. For all its recent revival, democratic self-defense scholarship carries two blind spots. The first concerns what we defend against: foreign state-level threats remain largely outside the frame. The major works of the past decade expressly confine themselves to internal threats, on the reasonable assumption that the hard question is justifying limits on home-grown actors who turn their own rights against democracy. A state’s defense against violent threats from abroad raises no comparable difficulty. Yet violence plays little part in foreign information manipulation, which operates covertly, woven into ordinary public discourse. It is thus no easy case: the manipulation is so tightly bound up with protected speech that countering it readily burdens the same liberal guarantees that make the domestic case fraught.

     The second concerns how we defend: even as scholars extend the field to digital challenges, most responses stay fixed on content—true versus false, harmful versus permissible—and so remain locked in a running tension with freedom of expression. Of late, however, a growing body of work has begun to argue that policing content is a reactive game of whack-a-mole, never reaching the systemic harm beneath it. Manipulation today runs on data-driven targeting, AI-generated content at scale, and algorithm-driven platform amplification. Even so, that insight has yet to be carried through to a fully structural response of the kind I develop below.

What a Democracy Must Defend

     The way past both begins with a more basic question: what, exactly, must a democracy defend? In the digital age, the answer is not this or that piece of content, nor any single right in isolation, but the integrity of its information environment—the structural conditions under which reliable information can be produced, circulated, and received. While freedom of expression and association are commonly cast as democracy’s floor requirements, I see the liberal rights of speech, privacy, and access to information as essential components of that condition, not the condition itself. Without the intermediary institutions through which information circulates, the rights alone cannot ensure that reliable information reaches the public. The two are mutually reinforcing. Understood this way, the integrity of the digital information environment is an institutional precondition of democratic self-government, significant enough to warrant active defense.

     Nor must the defense stop at the border: the foreign dimension was there in Karl Loewenstein’s 1937 essays from the start. The fascism he set out to resist was a transnational movement coming from authoritarian states. The “[p]arrying of subversive activities directed against the state from outside,” he wrote, constituted “one of the fundamental, and at the same time most subtle, functions of the democratic states.” Recovering this dimension is not a departure from the tradition. It is a return to its origins.

     That defense falls to the state. Individuals, and even civil society, cannot impose binding obligations on the infrastructure through which information now flows—and they certainly cannot reach a foreign regime. Where the threat is domestic, the state’s role can remain in the background, since its ordinary legal machinery can reach those responsible. Where it comes from a foreign state, the calculus changes: the regime enjoys sovereign immunity and sits beyond every domestic check; sanctioning its downstream proxies leaves it untouched. The state’s defensive role must move from the background to the foreground.

The Structural Turn

     But moving to the foreground raises the question of how. Here, then, is the reorientation: if the threat is structural, so must the defense be—targeting the manipulable infrastructural conditions through which information is produced, circulated, and received. Content regulation treats downstream symptoms while the manipulation operates upstream. That is precisely why practitioners, and the EU’s own framework on foreign information manipulation, long ago shifted their analysis from content to coordinated behavior and operations. The alternative of “more speech” fares no better: more speech is exactly what powerful state actors weaponize, flooding the environment to exhaust the scarce resource of public attention. Both horns of the content-versus-speech dilemma are dead ends.

     The way out is to govern the environment’s three infrastructural layers rather than its content: data protection and cross-border transfer safeguards, grounded in privacy law, for the personal data through which audiences are profiled and targeted; safety standards and misuse prevention for the AI systems increasingly embedded in how information is produced; and obligations on platforms to detect and mitigate coordinated inauthentic behavior, and to trace the covert foreign money behind it. At every layer, providers subject to authoritarian control should bear heightened obligations as a condition of market access.

     Regulating data, system safety, and behavior rather than what anyone says, these measures are largely content-neutral. They substantially ease, though do not eliminate, the tension with free expression that has long constrained democratic self-defense, ensuring that the state cannot set itself up as an arbiter of truth or silence dissent in the name of countering threats. Narrow rules for direct incitement and plainly illegal content still have their place, of course, and even content-neutral governance must be held to clear legal authority, due process, judicial review, and proportionality, lest state power escape its constitutional checks. But the claim stands: the center of gravity should shift from contested judgments about truth toward the structural conditions of the information environment.

Defending Without Destroying

     What to defend, against what, and how—behind every question this essay has asked stands the field’s founding paradox: how can a democracy defend itself without destroying itself in the process? The digital age has carried that paradox onto new terrain, where the prevailing answers remain too narrow about the threat and misplaced in their method. Defending the structural conditions of the digital information environment against foreign state threats offers, I believe, a better way through: maintaining what makes democratic self-government possible at all. For frontline democracies and beyond, the cases will keep coming. The theory should be ready to meet them not after the fact, but before.

You-Hao Lai, S.J.D. Candidate, The George Washington University Law School; LL.M., Harvard Law School (2023). Deputy Director, Democratic Governance Program, Research Institute for Democracy, Society, and Emerging Technology (DSET), Taiwan.

Suggested Citation: You-Hao Lai, ‘From Content to Structure in Democratic Self-Defense’, IACL-AIDC Blog (25/06/2026)