—Michalina Marcia, Postdoctoral Researcher at the Poznań Human Rights Centre, Institute of Law Studies of the Polish Academy of Sciences and at the Digital Justice Center, Faculty of Law, Administration and Economics, University of Wrocław, Poland
In recent years, technological influence on elections, facilitated by growing digitalisation, has become increasingly visible. From the deepfake incident during Slovakia’s 2023 elections, through heavily AI-influenced electoral campaigns in Argentina and Taiwan in 2023 and 2024, respectively, Germany’s 2025 federal elections, and the spread of distorted materials during the United States presidential elections, the examples prove that the threat is no longer speculative.
Elections are no longer shaped only in polling stations, campaign rallies, or televised debates. Increasingly, they are shaped in algorithmically curated digital environments, where voters encounter political reality through recommendation systems, targeted messages, synthetic media, and coordinated influence operations. It is becoming more and more difficult for citizens to exercise their electoral rights free from interference and manipulation. Even more concerningly, it is becoming increasingly difficult to detect such interference when it occurs. The technology-induced threats to electoral integrity vary in scale, visibility, and intrusiveness. They range from platform-based forms of content suppression or amplification, such as shadowbanning or opaque algorithmic prioritisation by online platforms, to deepfakes and microtargeted political advertising disseminated by candidates, parties, supporters, or foreign entities.
The nature of these threats raises pressing questions for contemporary constitutionalism. First, the digital character of the interference blurs the traditional line between internal and external threats to electoral processes, making it harder to identify the source of the threat and to design an adequate institutional response. Second, it raises the question of whether existing constitutional institutional frameworks are capable of responding effectively to challenges that now occur on a mass scale and at high speed. Third, it forces us to ask what new measures could be introduced, whether the state is institutionally and technologically prepared to implement them, and, what seems to be forgotten at times, whether such implementation would genuinely resolve the contemporary issues of electoral processes rather than amplify them.
These challenges have been noticed at the international level, with the EU’s European Democracy Shield seeking to strengthen the resilience of democratic institutions and elections against disinformation and the CoE Secretary General’s 2026 Annual Report placing democratic resilience at the centre of European security. With that, it becomes increasingly clear that we are facing the emergence of what might be called digital militant democracy – a phase in which democratic states increasingly seek to defend electoral self-government against digital, AI-enhanced manipulation, disinformation, and hybrid interference.
Yet, despite the increasingly sophisticated language of democratic resilience, national constitutional and electoral systems remain remarkably unprepared to determine when digital interference has become serious enough to justify state intervention, which institutions should make that determination, and what consequences should follow. Notwithstanding attempts to address key challenges, including the criminalisation of electoral deepfakes in several states, the regulations remain largely fragmented. Where existing regulation remains insufficient, and where regulatory responses are fragmented or underdeveloped, public authorities may lack adequate tools to respond to digital interference in electoral processes. This is precisely where the temptation emerges to rely directly on constitutional provisions, namely the right to free elections – just like in the notable case of the Romanian Constitutional Court (RCC) ruling of 6th December 2024, where the RCC annulled the first round of the presidential election, finding that coordinated, opaque digital campaigning had distorted the overall fairness and integrity of the vote.
The constitutionalisation of the digital electoral environment is both necessary and dangerous. Digital interference may undermine the conditions of free electoral choice, but treating such interference as a direct constitutional violation can empower courts to invalidate democratic outcomes on the basis of uncertain evidence and undefined thresholds. As the next elections are approaching amid threats of digital interference, the task here is not simply to recognise digital threats to free elections, but to decide how constitutional law should respond to them – through ad hoc judicial intervention, through clear regulatory thresholds, or through a combination of both.
What constitutes free elections?
The issue of electoral rights and freedoms in the context of disinformation, including disinformation linked to AI tools, has already appeared in the case law of the European Court of Human Rights under Article 3 of Protocol No. 1 to the European Convention on Human Rights (ECHR), but to a limited extent.
According to the Court’s case law, the concept of free elections is closely linked to the condition that citizens must be sufficiently informed about relevant electoral issues. Free elections function most effectively when voters understand the political choices before them. Actual knowledge of politics is a key element of citizenship and enables the full exercise of electoral rights. Yet the Court has not provided a definitive answer as to which digital conditions, applications of AI, or forms of online manipulation may affect the right to free elections, nor has it clearly identified the threshold at which state intervention becomes constitutionally or conventionally required.
It remains debatable whether this substantive understanding belongs within the concept of constitutional electoral law. On one hand, disinformation and manipulation of electoral processes long predate generative AI and the digitalisation of political campaigns. On the other hand, the possibilities, scale, speed, and opacity of contemporary digital interference arguably justify treating it as a distinct constitutional threat. The answer also depends heavily on the specific constitutional system and on the wording and interpretation of national constitutional provisions. Constitutional electoral laws have traditionally been concerned with preventing the state from interfering directly with voters’ freedom of choice, whether through legislation, executive action, or other forms of pressure.
Yet the right to vote, and the broader right to free elections protected by Article 3 of Protocol No. 1 ECHR, cannot be reduced to the formal act of casting a ballot. The very concept of “choosing” itself presupposes freedom and the ability to make an informed choice. In the context of the substantive content of electoral rights, the premise of informed choice, and the unprecedented digitalisation of political campaigns, it is therefore necessary to ask whether AI and digitally amplified disinformation can threaten free and informed participation in elections in the constitutional sense. This may be true even where such interference does not amount to a classical or formal violation of electoral rights. At stake may be the full realisation of the principles and objectives that constitutional electoral rights are meant to protect.
This broader interpretation was adopted by the RCC in its ruling of 6 December 2024, which annulled the electoral process and ordered it to be repeated due to massive Russian hybrid attacks carried out before the first round of voting. The RCC emphasised that voters’ rights include both the right to obtain accurate information about candidates and the electoral process from all sources, including online sources, and the right to protection against unjustified influence on voting behaviour based on unlawful and disproportionate actions. The RCC found that one candidate’s use of aggressive promotion through mass-media platform algorithms contributed to disinformation and manipulation of the vote, particularly because the campaign materials promoting that candidate did not bear the specific hallmarks of election advertising. It also found that the opaque use of digital technologies and artificial intelligence created a clear imbalance among candidates. In doing so, the RCC effectively broadened the interpretation of constitutional electoral provisions in response to the growing spread of disinformation and new threats to constitutional rights and freedoms. It suggested that electoral rights do not depend solely on formal electoral rules, but also on the broader environment, including the digital environment, in which citizens make political decisions, opening the scope of the right to free elections to new ways of understanding.
Digital militant democracy, court discretion and way forward
The general reaction to the RCC’s ruling was relatively positive. Many saw it as a proactive response to serious, fast-developing threats. Yet the adoption of such an ad hoc militant-democratic approach may also produce serious negative consequences. As the Romanian example shows, recognising digital electoral integrity as part of the constitutional right to free elections has far-reaching implications. It may justify annulling a vote, repeating elections, and directly intervening in electoral proceedings. In certain cases, it may even override or suspend the immediate expression of the popular will. This represents a modern form of militant democracy and, like all forms of militant democracy, it carries an unavoidable paradox: democracy may be defended by restricting or overriding rights and freedoms that are themselves central to democracy.
This creates several difficult questions about institutional responsibility and legal thresholds. It is not obvious, for example, who should determine that digital interference has crossed the line from ordinary campaign activity into a constitutionally relevant breach. Nor is it clear what scale of interference should matter – the number of voters reached, the volume of false or manipulated material circulated, or maybe the timing of the intervention. Interference during the official campaign may be treated differently from activity occurring months earlier, even if both shape electoral opinion. The same uncertainty arises in relation to scope. Should the rules apply only to online platforms, or also to encrypted messaging services, informal supporter networks, and foreign information operations? Finally, the most important question of all – what consequences should follow, and which institution should impose them?
If technological interference is treated as a direct infringement of the right to free elections and if courts, especially constitutional courts, are left to interpret and apply this standard ad hoc, the response system gains flexibility. It becomes capable of reacting to threats that cannot yet be foreseen and adapting to environment that changes quickly. But it also generates discretion, uncertainty, and institutional risk. In conditions of democratic erosion (that, as one may say, we are currently facing) opening such a door without clear limits may be dangerous. Tools that allow a court to protect democracy from manipulation may later be used to delegitimise elections, silence opposition, or produce selective constitutional interventions.
Of course, this does not mean that constitutional courts should play no role. On the contrary, courts remain essential guardians of fundamental rights and freedoms, and they must be able to respond to individual situations that cannot be fully predicted in advance. But now that the risks connected to digital disinformation and AI-enhanced electoral manipulation are visible, the better approach is not to leave the entire burden to judicial improvisation. The better approach is to create clear legal thresholds within a coherent regulatory framework.
Leaving courts to deal with digital threats to free elections without parallel regulatory efforts is institutionally irresponsible. Courts may lack the technical expertise, as well as the institutional and procedural capacity, to assess digital activity. They may struggle to distinguish organic political communication from coordinated manipulation, lawful persuasion from unlawful influence, or domestic political misconduct from foreign hybrid interference. They may also be vulnerable to instrumentalisation. For example, a party might circulate or amplify a deepfake concerning itself in order to provoke institutional chaos, delegitimise the electoral process, or force a constitutional court into intervention.
The challenge, therefore, is not whether constitutional law should respond to the digital transformation of elections. It must. The challenge is how to respond without converting the right to free elections into an open-ended instrument of post-electoral uncertainty. Given the number of examples of electoral interference, it cannot be stressed enough – whether we determine that a manipulation-free digital environment is an integral part of free elections or not, we are at the point of urgent need for clear regulation that includes technologically facilitated disinformation in electoral law provisions rather than relying only on individual court decisions. To support the criminal law provisions, labelling requirements and online platform oversight, we also need inclusion of unlawful digital interferences into the election-related rectification and electoral objection procedures in order to react quickly and to prevent the disinformation from spreading and influencing the electoral processes. These should include entities entitled to initiate the procedures, reasons that define technical interference and competent bodies. Finally, there must be clear paths for transinstitutional cooperation between courts, electoral bodies and digitally-specialised, cybersecurity agencies as a way for guaranteeing technically valid decisions, not based on a “general impression”.
The digital environment should be recognised as part of the constitutional architecture of elections, but that recognition must be accompanied by clear standards, institutional safeguards, and procedural discipline. Otherwise, we may solve one democratic problem by creating another.
Suggested citation: Michalina Marcia, Digital Militant Democracy: Free Elections, Technological Interference and the Limits of Constitutional Self-Defence, Int’l J. Const. L. Blog, Aug. 1, 2026, at: http://www.iconnectblog.com/digital-militant-democracy-free-elections-technological-interference-and-the-limits-of-constitutional-self-defence/