Cognitive warfare (CW) exploits vulnerabilities in democratic media, institutional and civic infrastructures. This study asks how internal structural decay—specifically market-driven reforms, populist governance, and regulatory capture—creates openings for CW operations. A comparative analysis reveals that vulnerability stems from these intersecting domestic factors rather than from democratic openness. Adversaries leverage multi-level feedback loops of disinformation and polarisation to degrade epistemic foundations. Effective responses require systemic reforms: reinvestment in public-interest media, robust institutional safeguards, and enhanced civic education— to rebuild societies’ capacity for collective sense-making and resilience.

From Epistemic Erosion to Cognitive Conflict: A Multi-Level Analysis of Democratic Vulnerability

KP Kiely¹, Y Naydenova²

¹GATE (Big Data for Smart Society) Institute, Sofia University

Sofia, Bulgaria

E-mail: Keith.Kiely@gate-ai.eu

² GATE (Big Data for Smart Society) Institute, Sofia University

Sofia, Bulgaria

E-mail: Yana.Naydenova@gate-ai.eu

Abstract: Cognitive warfare (CW) exploits vulnerabilities in democratic media, institutional, and civic infrastructures. This study asks how internal structural decay—specifically market-driven reforms, populist governance, and regulatory capture—creates openings for CW operations. A comparative analysis reveals that vulnerability stems from these intersecting domestic factors rather than democratic openness. Adversaries leverage multi-level feedback loops of disinformation and polarisation to degrade epistemic foundations. Effective responses require systemic reforms: reinvestment in public-interest media, robust institutional safeguards, and enhanced civic education to rebuild societies’ capacity for collective sense-making and resilience.

Keywords: Cognitive Warfare; Democratic Vulnerability; Constructivism; Disinformation; Epistemic Infrastructure.

Introduction

Cognitive warfare (CW) has emerged as a central challenge for contemporary democracies. These operations target not just information systems but also trust, shared meaning, and institutional legitimacy through coordinated campaigns of false information and algorithmic manipulation (Zuboff 2019; Benkler et al. 2018). The evolution of artificial intelligence (AI) and digital media ecosystems has deepened the fragmentation of epistemic foundations, making open societies innately vulnerable (Adler-Nissen & Pouliot 2014; Brown 2015). For military, intelligence, and security communities, this transforms the information environment into a durable operational battlespace in which structural weaknesses in democratic societies become critical targets and constraints for information and cognitive operations. This paper asks: How do internal structural decays in democratic societies—specifically within media, institutional, and civic infrastructures—create exploitable vulnerabilities for cognitive warfare (CW) operations? While democratic openness enables pluralism, structural vulnerabilities increasingly result from policy choices and social change: market-driven reforms, populist pressures, and crisis governance have systematically weakened epistemic infrastructure and institutional safeguards (Crouch 2004; Abernathy 2020). A combined array of factors including but not limited to media deregulation, the decline of local journalism, and the rise of algorithmic curation continue to create fertile ground for foreign actors or local fringe groups to exploit (Sunstein 2017; Couldry & Mejias 2019). These patterns are not geographically limited to Europe or North America visible across diverse systems self-identified as democracies, from Europe to Latin America. A systematic assessment of how cracks in societal foundations contribute to susceptibility is therefore not only analytically overdue but strategically necessary for planning and defending against contemporary information and cognitive operations.

Yet, despite widespread concern about CW, policy and academic debates continue to focus primarily on state-led foreign interference. This narrow emphasis risks underestimating how domestic meaning-making, institutional weaknesses, politicized media, endemic corruption, and socio-economic inequality create the conditions for both foreign and homegrown manipulation. Neglecting institutional and structural factors ultimately obscures the enabling environment in which misleading or manipulative information can circulate and gain traction (McChesney 2019). More effective responses require moving beyond rhetorical signalling in national and international arenas towards addressing these internal vulnerabilities directly. This includes policies to rebuild and modernize epistemic infrastructures, reinforce institutional safeguards, and strengthen civic agency through sustained investment in critical and civic educations (Habermas 1984). Such measures can substantially reduce the space in which FIMI campaigns can operate successfully (OECD 2023). This article responds to that gap by offering a structural vulnerability framework and heuristic composite index that enable analysts and practitioners to integrate media, institutional, and civic risk factors into assessments of contested information spaces across multiple democracies.

This study develops a comparative multi-level analysis of democratic vulnerability to cognitive warfare across a set of Western, Central/Eastern European, and non-Western democracies. Methodologically, it introduces a heuristic composite index that summarizes existing indicators of institutional, media, and civic fragility to visualise how structural risk factors cluster and create exploitable conditions for information and cognitive operations. The index is explicitly exploratory: it does not validate causal claims, but visualizes how structural risk factors cluster and align with documented episodes of cognitive warfare and Foreign Information Manipulation and Interference (FIMI). 'Epistemic infrastructure' here refers to the institutions and norms that enable societies to produce and validate shared knowledge; 'cognitive autonomy' means the capacity for critical judgment and resistance to manipulation. Our analysis draws on secondary literature, NGO reports, and global datasets; further methodological detail is provided in the Methods section.

The following sections will proceed as follows, first we conceptualize the mechanisms of cognitive warfare, then map structural vulnerabilities across contexts, present a multi-level framework for systemic impact, and finally discuss general policy implications for democratic resilience.

Literature Review

Definitions and Evolution: Cognitive Warfare: Beyond Information Warfare

Cognitive warfare (CW) represents a qualitative shift from traditional Information Warfare (IW), targeting not merely the information environment but the interpretive processes through which individuals and collectives construct meaning and make sense of their social reality. While IW focuses on controlling, denying, or manipulating information flows within contested environments—targeting what people know—CW weaponises the cognitive and interpretive frameworks themselves, exploiting how people process, evaluate, and internalize information (Deppe & Schaal, 2024; NATO Science & Technology 2022).

This distinction is operationally critical. Traditional IW campaigns seek to establish information dominance: controlling narratives, suppressing counter-messages, and shaping the informational landscape to constrain adversary decision-making (Arquilla & Ronfeldt 1996). CW, by contrast, targets epistemic infrastructure—the institutions, norms, and cognitive processes that enable societies to distinguish fact from fiction, signal from noise. Rather than simply controlling information, CW degrades societies' capacity for collective sense-making, eroding the shared epistemic foundations upon which democratic deliberation depends (Miller 2023; NATO Allied Command Transformation 2021).

From a constructivist perspective, CW exploits the socially constructed nature of knowledge production itself. Where IW treats information as a resource to be controlled or denied, CW recognizes that meaning emerges through intersubjective processes—shaped by trust in institutions, shared interpretive frameworks, and collective cognitive habits (Wendt 1992; Hopf 2013). By systematically undermining confidence in journalism, science, electoral institutions, and other epistemic authorities, CW creates what can be termed epistemic orphanhood: a condition in which citizens lack credible reference points for evaluating truth-claims and retreat into tribalized information ecosystems or disengagement (Wright 2025).

The integration of digital technologies and generative AI has amplified CW's operational scope. Platform algorithms curate individualized realities through microtargeting and engagement optimization, fragmenting the public sphere into self-reinforcing echo chambers (Zuboff 2019; DiResta et al. 2021). Large language models enable the mass production of hyper-personalized, emotionally resonant disinformation that can bypass human-scale fact-checking and exploit cognitive biases at unprecedented speed and granularity (Williams et al. 2024; Canadian Centre for Cyber Security 2023). These technologies do not merely accelerate information manipulation; they transform cognition itself into a targetable domain by weaponising the interpretive process—the mechanisms through which individuals form beliefs, assign trust, and construct political identities.

This shift has profound implications for democratic vulnerability. While democracies have long contended with propaganda and disinformation, CW operates at a deeper structural level: degrading the epistemic infrastructure and cognitive autonomy necessary for self-governance. The challenge is no longer simply to counter false narratives but to rebuild the intersubjective foundations—institutional trust, critical capacity, and shared reality—that enable democratic publics to resist manipulation and engage in meaningful collective deliberation.

Current Approaches: The Dominance of State-Led FIMI in the Literature

Within the broader field of information warfare and information operations, a substantial amount policy and academic work treats “cognitive warfare” as an extension of state‑led foreign information manipulation and interference (FIMI), emphasising large‑scale social media campaigns, cyber tactics, artificial amplification, and AI‑driven targeting that typically remain below the threshold of conventional conflict (Bradshaw & Howard 2019; Deppe & Schaal 2024). In this perspective, the primary threat is exogenous: hostile states or organised networks are seen as attempting to penetrate otherwise bounded and largely coherent democratic publics (Kuo & Marwick 2021; Snyder 2018; Splidsbøel‑Hansen 2021).

This focus has shaped a corresponding policy repertoire. European Union (EU) and North Atlantic Treaty Organisation (NATO) frameworks integrate FIMI into hybrid‑threat doctrines and prioritise technical and tactical responses such as fact‑checking, platform cooperation, cybersecurity hardening, and monitoring of information operations (Bradshaw & Howard 2019; Deppe & Schaal 2024). While these measures mitigate specific campaigns, they largely leave untouched the deeper epistemic and institutional conditions that make such campaigns effective. In other words, they respond to hostile messages and manipulative operations without systematically addressing how media deregulation, institutional decay, and the erosion of civic and critical education have already weakened epistemic infrastructures and cognitive autonomy (Brown 2015; Benkler, Faris & Roberts 2018).

This article advances a different starting point. Rather than treating CW solely as a foreign or exogenous phenomenon, it foregrounds domestic structural vulnerabilities as integral to the contested space. By focusing on how long‑term policy choices, market logics, and governance practices have reshaped media systems, institutions, and democratic agency, the analysis reconceptualises CW as operating within—and through—pre‑existing fractures in democratic epistemology. The remainder of the paper therefore shifts from cataloguing external information operations to developing a structural vulnerability framework and heuristic index that make these domestic enabling conditions analytically visible and empirically comparable across democracies.

Theoretical Framework

A Constructivist Lens on Structural Vulnerability

Wendt’s constructivist framework provides a critical lens for understanding the structural vulnerabilities that cognitive warfare (CW) exploits in contemporary democracies (Wendt, 1992). Structures are not static or purely material. Instead, they are socially constructed through ideas, norms, and shared practices (Adler 1997; Blyth 2002). Within this framework, power and institutions derive legitimacy from intersubjective understandings reproduced through discourse and policy (Schmidt 2008; Onuf 2012). Therefore, market reforms, populist pressures, and crisis governance are not exogenous shocks, but products of changing collective interpretations of legitimacy and authority.

This perspective reframes phenomena like media deregulation, institutional decay, or eroding civic agency as co-constructed outcomes of contested and fluidic social narratives. For instance, the normalization of ‘patriotic journalism’ in Hungary and reforms targeting critical citizenship in the United States demonstrate how agency and vulnerability are both constructed and reversible (Hornbeck & Malin, 2023). Because vulnerabilities are socially generated, they are also open to deliberate change through reform and civic engagement.

Contemporary constructivist scholarship extends this reasoning, treating CW as a struggle over the social construction of democratic reality itself. Hopf (2013), for example, shows how CW weaponises identity and memory to fragment societies, while other theorists describe how norms, data and information infrastructures become targets for manipulation or contestation (Couldry & Mejias 2019; Bigo, Bonditti, & Olsson 2011). Laclau and Mouffe’s (1985) discourse theory complements this by illustrating how those spreading false or misleading information exploit the contingency of political identities, rearticulating democratic narratives through “empty signifiers” to create hegemonic configurations that can become hostile to pluralism.

In sum, these perspectives frame CW as an ideational struggle over the meaning and boundaries of democracy. Democratic resilience depends on rebuilding intersubjective infrastructures—networks of trust, legitimacy, and critical capacity—via counter-narratives, policies which prioritise public good, institutional transparency, and reviving civic agency. Ultimately, defending democracy in the cognitive domain requires recognizing that power lies not just in institutions or technology, but in society’s ongoing process of meaning-making (Bigo, Bonditti, & Olsson 2011).

Cognitive Warfare Mechanisms: Targeting Epistemic Infrastructure and Autonomy

Cognitive warfare (CW) marks a shift in conflict dynamics: its objective is not only to manipulate information, but to degrade epistemic infrastructure—the institutions, norms, and technologies that allow societies to produce, validate, and act on shared knowledge (Miller 2023). From a constructivist perspective, these infrastructures are sustained by intersubjective trust, shared meanings, and collective practices (NATO 2022). CW exploits this by targeting three core mechanisms.

First, epistemic trust erosion occurs when bad actors or internal institutions systematically undermine confidence in key nodes of democratic epistemology—journalism, science, elections—through campaigns that weaponise scepticism, ridicule, and information overload (Wright 2025). Citizens, deprived of credible authorities, experience a kind of epistemic orphanhood, retreating into tribalized realities or withdrawal.

Second, cognitive autonomy subversion floods the public sphere with disorienting narratives and installs pre-reflective interpretive defaults that displace deliberation with reactive scripts (Kahneman 2011). his dynamic echoes Arendt’s (1967) warning mentioned early that when lies become indistinguishable from facts, politics devolves into competing fictions rather than meaningful policy debate. Manipulation thus operates through the social construction of meaning and collective cognitive habits, not just through false content.

Third, algorithmic reality curation leverages AI-driven microtargeting and engagement algorithms to create self-reinforcing epistemic bubbles (Zuboff 2019). As Haigh and Andrusenko (2021) show, attacks on signal validation layers can corrupt decisions at the sensor level—an analogous vulnerability arises in human cognition when perceived trustworthiness trumps verification. These algorithmic practices actively shape what is visible, plausible, and legitimate in public life, narrowing the horizons of the thinkable (DiResta et al. 2021).

In operational terms, these mechanisms map onto observable tactics: attacking institutional credibility and election integrity (epistemic trust erosion), deploying emotionally charged and identity-based narratives that short-circuit deliberation (cognitive autonomy subversion), and exploiting platform algorithms, bots, and microtargeting to curate closed information bubbles (algorithmic reality curation).

Democratic Vulnerability as Structural: The Neoliberal-Epistemic Nexus

The susceptibility of democracies to cognitive warfare is amplified by structural transformations that weaponise democratic ideals against their own foundations. Vulnerabilities are shaped by market-oriented policies, populist pressures, and crisis-driven governance (Brown 2015; Crouch 2004). Collectively, these factors erode the institutions and practices that sustain rational-critical discourse (Acemoglu & Robinson 2021; Benkler, Faris & Roberts 2018).

The paradox of openness is that democratic institutions—free press, independent judiciary, academic peer review—rely on distributed trust systems to sustain public deliberation (Habermas 1984). Since the late 1970s and 1980s, neoliberal orientations have progressively weakened these systems. Media deregulation and the collapse of local journalism have produced “news deserts” increasingly filled by algorithmically amplified disinformation (Gandini et al. 2025; Abernathy 2020). Austerity-driven defunding of public media and the privatization of expertise degrade epistemic gatekeeping (McChesney 2015). A public sphere dominated by engagement-driven platform algorithms tends to privilege emotionally charged content over accuracy, fracturing shared epistemic baselines and reinforcing polarised echo chambers (Zuboff 2019). Democratic openness thus becomes a vector for epistemic collapse when institutional safeguards are dismantled (Welch 2013).

This fragility is reinforced through three interconnected pathways (Zuboff 2019). First, the epistemic–industrial complex describes market-driven media ecosystems that outsource truth-production to attention-maximizing platforms, generating economies of scale for disinformation (European External Action Service 2022). Second, democratic agency erosion—linked to precarious labour conditions (Standing 2011) and defunded civic and humanities education and defunded civics education (Iyengar et al. 2019) weakens critical engagement and heightens susceptibility to affective polarisation (Persily 2020). Third, institutional capture (IC) involves revolving-door politicization and regulatory stagnation, leaving electoral and digital infrastructures vulnerable to exploitation under the guise of transparency and openness (Brown 2015). Empirical cases illustrate how these dynamics extend beyond classic neoliberal contexts. In Brazil and the Philippines, populist leaders have weakened institutional safeguards and promoted media capture, increasing susceptibility to cognitive warfare even where neoliberal reforms are partial or contested (Novais & Leite 2024; Human Rights Watch 2022). Additionally, in both Brazil and the Philippines, dense integration of partisan politics with platform-mediated communication has effectively turned major social media ecosystems into persistent operational theatres where domestic and foreign actors can run   influence   campaigns   at   scale   with   limited   institutional   friction. As the European External Action Service notes, CW functions as a hybrid threat that exploits such democratic vulnerabilities to erode the perceived legitimacy of competing governance models (European External Action Service 2022).

The Large Language Model (LLM) Revolution: Hyper Personalised Cognitive Warfare

Building upon the structural transformations of the neoliberal-epistemic nexus, the Large Language Model (LLM) Revolution introduces a potent force multiplier by enabling hyper-personalised cognitive warfare. Recent empirical research demonstrates the alarming effectiveness of LLMs in generating convincing disinformation content. The DisElect study found that 13 tested LLMs could consistently produce election disinformation indistinguishable from human-written content over 50% of the time, with multiple models achieving "above-human levels of humanness" (Williams et al. 2024) Complementary research reveals that LLMs can generate compelling false news articles across 20 different disinformation narratives, with detection models struggling to identify AI-generated content (Canadian Centre for Cybersecurity 2023). Critically, these capabilities extend beyond text generation to include real-time sentiment analysis, enabling AI systems to analyse facial expressions, vocal tones, and written content to pinpoint emotional vulnerabilities and tailor manipulative messages accordingly (Shi et al. 2024). Large language models and related AI systems further reduce the cost, increase the speed, and expand the granularity with which adversaries can generate, tailor, and test disinformation and influence narratives against structurally vulnerable publics. Emotional prompting emerges as a particularly concerning vector for exploitation. Research indicates that LLMs show significantly higher disinformation production rates when prompted with polite language compared to neutral requests, suggesting that social engineering techniques can bypass safety mechanisms (Shi et al. 2024). This vulnerability is compounded by inconsistent safeguarding across models: while effective safeguards are technically feasible, they are "inconsistently implemented" across different LLMs, creating exploitable gaps.

The implications for democratic epistemology are profound. LLMs enable the mass production of subtly varied propaganda that can build trust through mimicking human interaction while integrating disinformation into online communities (Santana 2025). This capability transforms cognitive warfare from a resource-intensive operation requiring human expertise into an automated, scalable process that can generate "high-quality content for election disinformation operations, even in hyper localised scenarios, at far lower costs than traditional methods (Williams et al. 2024). The resulting fragmentation of shared epistemic baselines occurs not through obvious manipulation but through the gradual erosion of the distinction between authentic and synthetic discourse.

From a constructivist perspective, LLMs represent a fundamental challenge to ideas around the intersubjective foundations of democratic knowledge production. The personalization capabilities of LLMs exploit the very cognitive processes through which individuals form beliefs and make democratic choices, contributing to transforming openness from a democratic strength into a systematic and targetable vulnerability. In the structural framework developed here, AI‑enabled cognitive operations are therefore not a separate threat vector but a force multiplier: they interact with pre‑existing vulnerabilities in media, institutions, and civic capacity, amplifying the risks captured by the composite vulnerability scores.

Research Design

Comparative Design and Rationale

This study employs a structured comparative analysis using a Most-Different Systems Design (MDSD) across three regional clusters to examine how structural vulnerabilities enable CW

across diverse democratic contexts (Przeworski & Teune 2024). This design moves beyond the existing literature's focus on state-centric Foreign Information Manipulation and Interference (FIMI) by systematically analysing domestic structural conditions that amplify CW susceptibility. The logic of MDSD allows for the identification of common causal factors by comparing cases that are dissimilar in many respects but share a common outcome—in this instance, susceptibility to cognitive warfare (Lijphart 1971). The composite index outlined below is used as a descriptive device to integrate established indicators into a single comparative lens. The weighting scheme is normative and theory‑driven rather than statistically validated, and results should be interpreted as heuristic patterns, not as precise causal estimates or rankings. No formal causal inference (example, regression, matching, or panel designs) is attempted.

Case Selection Framework

Cases were selected using a five-dimensional matrix ensuring variation across key structural vulnerability factors while maintaining analytical coherence. The empirical analysis covers twelve democracies: ten structurally vulnerable cases (United States, Sweden, Greece, Hungary, Poland, Bulgaria, Romania, Brazil, India, and the Philippines) and two resilient comparison cases (Switzerland and Finland). The study employs a diverse‑cases strategy, selecting ten structurally vulnerable democracies plus two resilient cases across Western Europe/North America, Central and Eastern Europe, and the Global South to capture pronounced variation in media structures, institutional quality, and civic conditions while holding the basic commitment to electoral competition and formal democratic institutions constant.

Case selection follows two criteria. First, a comparative variation logic: the ten structurally vulnerable cases (United States, Sweden, Greece, Hungary, Poland, Bulgaria, Romania, Brazil, India, and the Philippines) are chosen to represent distinct regional clusters and institutional models, maximizing variation in media regimes, institutional quality, and civic conditions under a shared democratic umbrella. Second, theoretical relevance: each country has featured prominently in research or NGO reporting on disinformation, foreign information manipulation and interference, or media capture, and thus provides empirically salient instances of the mechanisms of epistemic trust erosion, cognitive autonomy subversion, and algorithmic reality curation that are central to the framework. The sample is therefore designed to be illustrative and heuristic, not statistically representative

Operationalising the Constructivist Framework

To bridge the theoretical framework with our empirical analysis, we treat the 0–3 coding scales not merely as counts of material failures, but as measures of intersubjective and normative stability. In line with the Constructivist lens established in Section 3.1, the Epistemic Infrastructure Decay (EID) score measures the degree to which shared knowledge production has shifted from a public-interest norm to a market-driven or politicised practice. Institutional Capture (IC) is coded as the discursive and legal erosion of regulatory independence, reflecting the "revolving door" dynamics that redefine institutional meaning. Finally, Democratic Agency Erosion (DAE) quantifies the structural loss of critical capacity, where citizenship transitions from an active practice of self-governance to a passive mode of datafied consumption. By mapping the mechanisms of epistemic trust erosion and algorithmic curation onto these tiers, the heuristic index captures the degree to which a society’s "intersubjective infrastructures" have been hollowed out, leaving them susceptible to narrative and cognitive warfare.

Enhanced Data Sources and Case Selection

To ensure empirical robustness and address potential source biases, this study systematically triangulates evidence from a diverse set of quantitative and qualitative sources:

• Freedom House media freedom scores for temporal variation (Freedom House 2024).
• Transparency International corruption indices to measure institutional capture (Transparency International 2024).
• Reuters Institute Digital News Reports for media consumption patterns (Newman, Fletcher, Robertson, & Nielsen 2024).
• World Bank Worldwide Governance Indicators for regulatory quality (World Bank 2024).

These datasets are harmonized and extracted for each case across four temporal periods (baseline, transition, vulnerability, impact), facilitating both cross-sectional and longitudinal analysis. These datasets are selected as proxies for the health of social practices; for example, the V-Dem (Varieties of Democracy) Liberal Democracy Index is used to measure the 'reproduction of legitimacy' and the stability of the shared epistemic base established in our theoretical framework. To test the explanatory power of the structural vulnerability framework, the research design includes two negative and resilient cases—specifically Switzerland, and Finland. These cases are selected due to their combination of market economies and high democratic quality, yet relative immunity to large-scale cognitive warfare. The analysis examines both periods of attempted interference and episodes of demonstrated resilience, identifying structural and policy factors that inhibit manipulation.

All cases, including negative and resilient examples, are coded using the standardized vulnerability framework, enabling systematic comparison and the identification of both risk and protective factors across diverse democratic contexts.

Structural Vulnerability Coding Scheme

The analysis employs a three-tier coding framework measuring:

Epistemic Infrastructure Decay (EID): Erosion of institutions enabling shared knowledge production (media concentration, local journalism collapse, platform algorithm dominance)

Institutional Capture (IC): Regulatory compromise (revolving door dynamics, media regulator politicization, anti-corruption mechanism weakening)

Democratic Agency Erosion (DAE): Decline in citizens’ capacity for critical engagement (economic precarity, civic education defunding, digital manipulation exposure)

Each category uses standardized 0–3 scales with contextual modifiers for Cultural-Identity Fragmentation (CIF) and Crisis Governance Patterns (CGP). The composite vulnerability score is calculated as:

This enables cross-case comparison and quantitative analysis of structural vulnerability patterns. For practitioners, this composite score is intended as a prioritisation and situational-awareness tool: it highlights where structural conditions create a more favourable or hostile environment for the planning and effectiveness of information and cognitive operations,    rather    than    predicting    specific    campaign    outcomes. At the upper end of the scale, a composite score approaching 3 signifies a state of “epistemic orphanhood,” in which distinctions between authentic and synthetic discourse have largely dissolved and shared orientation to reality has collapsed, creating a highly permissive environment for cognitive operations. The Cultural-Identity Fragmentation (CIF) and Crisis Governance Patterns (CGP) modifiers are essential for capturing the "intersubjective understandings" and "identities" that those seeking to manipulate and weaponise.

Data Collection and Triangulation Strategy

Data collection employs multi-source triangulation combining:

Quantitative indices: V-Dem (V-Dem Institute 2024), Freedom House (2024), Transparency International (2024), Reuters Institute (2024), World Bank (2024).

Policy documentation: Legislative databases, government transparency portals Media analysis: Euromedia Ownership Monitor (2023), national media regulators The Restructuring of Democratic Epistemology

Media Deregulation and Epistemic Crisis

The structural erosion of democratic epistemology begins with media deregulation, privatization, and consolidation, which have weakened traditional gatekeeping and amplified societies’ vulnerability to cognitive warfare (CW). The 1996 U.S. Telecommunications Act, for example, enabled unprecedented media consolidation that drastically reduced the number of independent local news outlets. This policy-driven process created widespread "news deserts," leaving entire communities with limited access to reliable, locally relevant information and dependent on national or algorithmically curated sources (Abernathy, 2020).

This causal pathway—from deregulation to epistemic vacuum—was vividly illustrated in the 2016 U.S. presidential election. Adversarial actors exploited these information-poor environments by micro-targeting residents with disinformation campaigns, leveraging social media platforms to amplify polarising and misleading narratives (Martin & McCrain 2019). This process demonstrates how structural changes in media ownership, driven by a constructivist shift in norms prioritizing market logics over public service, directly create vulnerabilities for CW (Blyth 2002). Parallel trends of media consolidation and the rise of oligarchic control over information are evident in Greece (Human Rights Watch 2025), Sweden (Garz & Ots 2025), and India (Shinde 2024), where declining journalistic pluralism has weakened democratic accountability (Kazaki 2022). The global result is a fragmented public sphere where platform algorithms prioritize engagement over accuracy, fracturing shared epistemic baselines into polarized echo chambers (Zuboff 2019). In the context of information warfare, this fragmentation significantly lowers the cost and increases the impact of targeted disinformation and influence operations, as bad actors can exploit existing news deserts and algorithmically segmented audiences rather than building fractures from scratch.

Institutional Decay: Regulatory Capture and Weakened Safeguards

The second structural driver, institutional decay, further weakens democratic resilience. Regulatory capture, a process theorized by Stigler (1971), has intensified in the digital era, with "revolving door" dynamics between government oversight bodies and regulated industries (The Revolving Door Project 2024) becoming commonplace in the U.S., UK (Brooks 2021), and Canada (TekSavvy Solutions Inc 2023). At the EU level, this phenomenon is also widespread, particularly in Central and Eastern European member states (Transparency International, 2020) where politically connected individuals are often appointed to media regulation bodies, resulting in weak enforcement and regulatory forbearance (European Audiovisual Observatory 2019).

In Hungary and Poland, (arguably in the United States) this process has been elevated to a deliberate strategy of institutional capture (IC), where ruling parties have systematically politicized public broadcasters and media regulators (V-Dem Institute 2024). From a constructivist perspective, this suggests a specific causal pathway. By redefining critical journalism as 'unpatriotic,' populist leaders constructed new social realities that justified institutional decay (Schmidt 2008). This captured institutional landscape, in turn, facilitates CW by eroding the very safeguards intended to protect democratic deliberation and electoral integrity (International Press Institute 2022). For hostile information actors, such captured environments offer dual advantages: they restrict independent counter-speech and fact-checking, while providing domestically produced narratives that can be amplified or repurposed in cross-border cognitive operations. Similar patterns of politicized oversight are evident in Bulgaria (Centre for Media Pluralism and Media Freedom 2024) and Romania (EU Disinfolab 2023), where low regulatory independence correlates with high susceptibility to disinformation.

Erosion of Democratic Agency: Precarity, Datafication, and Defunded Civic Education

Finally, democratic agency itself has been systematically eroded by three interconnected trends. First, rising economic precarity (Standing 2011), with over 40% of EU workers lacking stable employment, depresses formal political participation (Zhirnov 2024) and increases susceptibility to populist, anti-system narratives. Second, the datafication of political engagement—exemplified by the Cambridge Analytica scandal (Persily 2020)—has commodified civic life, replacing mass persuasion with individualized microtargeting that fragments the public sphere and erodes autonomous decision-making (Zuboff 2019). Third, the defunding of civic and humanities education (Giroux 2014) has left citizens ill-equipped to navigate complex information environments and resist digital manipulation (OECD 2020).

These trends illustrate another constructivist causal pathway: when civic education is deprioritized, it reflects and reinforces a shift in the shared meaning of citizenship, from an active practice of self-governance to a passive mode of consumption (Wendt 1996). This erosion of critical capacity creates fertile ground for CW, as individuals become more vulnerable to disinformation and disengagement (Wright 2006). Initiatives in Porto Alegre, Brazil (Satterthwaite et al. 2018), and elsewhere show that empowering citizens through participatory forums can help rebuild this eroded agency and foster collective resilience (Evans et al. 2023).

A Multi-level Framework of Analysis

To grasp the systemic impact of cognitive warfare (CW), it is necessary to move beyond a purely structural diagnosis and examine how these vulnerabilities manifest across multiple, interacting levels of democratic society. As the preceding analysis demonstrated, media deregulation, institutional decay, and eroded civic education create a cycle of epistemic fragility. CW's potency, however, lies in its ability to exploit these gaps simultaneously at the international, national, and individual levels (NATO Allied Command Transformation 2023). These tiers are deeply interdependent, creating a cascading effect: international destabilization weakens national resilience, while domestic polarisation and individual disengagement further undermine collective defences and social cohesion. The recent emergence of Large Language Models (LLMs) acts as a potent force multiplier across all three levels, scaling and intensifying these manipulations by enabling hyper-personalized disinformation and the automated curation of synthetic realities.

This multi-level approach, informed by constructivist theory, reveals that CW's efficacy stems not merely from exploiting weaknesses but from weaponising socially constructed realities, turning shared norms and identities into contested battlegrounds (Wendt 1992). These vulnerabilities are sustained through mutually constitutive feedback loops, where, for example, polarisation at the individual level reinforces populist national narratives, which in turn embolden adversarial actors internationally (Deppe & Schaal 2024). The following analysis briefly maps how bad actors leverage these interconnected domains, illustrating the full spectrum of risks facing contemporary democracies (Centre for the Study of Democracy 2025).

International, National, and Individual Levels of Analysis

At the international level, CW seeks to undermine the legitimacy and cohesion of democratic alliances. Russian disinformation campaigns during the Brexit referendum (Atlantic Council 2018), for instance, amplified anti-EU sentiment by framing the European Union as undemocratic and economically harmful, contributing to a climate of mistrust (Benkler, Faris & Roberts 2018). In Central and Eastern Europe, similar tactics have been deployed to fracture transatlantic solidarity by instrumentalizing historical grievances and ethnic divisions in the Baltic states (Belfer Centre for Science and International Affairs 2019), or by linking pro-Kremlin candidates to narratives of EU-driven cultural decline in Slovakia (Zgut-Przybylska 2024).

At the national level, CW exploits and amplifies polarisation, erodes institutional trust, and induces policy paralysis (Modern Diplomacy 2025). In the United States, foreign and domestic

actors have leveraged media fragmentation to deepen partisan divides, rendering consensus on democratic reforms nearly impossible (Persily 2017). In Central and Eastern Europe, this dynamic is more pronounced. In Hungary, the Orbán government’s systemic media capture has created a pro-government ecosystem that stifles dissent (Bajomi-Lázár 2024), while in Poland and Bulgaria, the politicization of regulatory bodies has eroded public trust (CMPF 2016). In Brazil, the Bolsonaro administration's attacks on the judiciary and media similarly paralyzed policymaking and eroded institutional legitimacy, illustrating a global vulnerability (Novais 2024). At the individual level, CW exploits universal cognitive biases through digital technologies. Social media algorithms engineer engagement by creating self-reinforcing echo chambers (Zuboff 2019), while the rise of generative AI enables hyper-personalized misinformation campaigns that are persuasive and difficult to detect (Sophos 2024). In the Baltics, Russian information operations have targeted ethnic minorities with tailored narratives to deepen social divisions (Belfer Centre for Science and International Affairs 2019), while in South Africa, WhatsApp-based disinformation during elections has exploited cognitive biases to fragment the information environment, particularly among rural voters (Allen & le Roux 2024).

Synthesis: Interdependence and Feedback Loops

The interdependence of these levels creates a cascading system of democratic decline. As the cases of Brexit, the 2016 U.S. election, Chinese influence operations in Taiwan, and Russian CW in the Baltics illustrate, international attacks on alliances weaken national institutions; national polarisation makes societies more receptive to divisive narratives; and individual-level epistemic fragmentation ensures that bottom-up pressure for anti-democratic policies persists. This multi-level framework suggests that CW is a systemic phenomenon, with each level acting as both a target and a vector for further disruption. The feedback loops identified here underscore the need for multi-dimensional, context-sensitive interventions to build genuine democratic resilience.

Table 1:Comparison- Multi-Level Cognitive Warfare Campaigns

The feedback loops identified here underscore the need for multi-dimensional, context-sensitive interventions to support and encourage genuine and durable democratic resilience.

Discussion, Recommendations, Conclusions

This study has shown that cognitive warfare (CW) is a structural threat that exploits the foundations of contemporary democracies. By targeting epistemic infrastructure, institutional trust, and civic agency, CW weaponises democratic openness—pluralism, transparency, and deliberation—against itself. Our multi-level framework reveals how CW operates in a cascading manner (Pamment & Bjola 2020) across international, national, and individual domains, creating feedback loops that amplify democratic decay (Levitsky & Ziblatt 2018). The cumulative results of this multi-indicator analysis are presented in Table 2. The low vulnerability scores for Finland (1.30) and Switzerland (1.67) serve as more than just statistical anchors; they provide empirical validation for the Constructivist claim that democratic susceptibility is not an inherent flaw of openness, but a socially constructed and reversible condition. These states demonstrate that resilience is a product of deliberate policy choices. For example, Finland’s long-standing investment in comprehensive media literacy and the protection of public-interest journalism has proactively built a shared epistemic foundation that resists fragmentation. This confirms that while vulnerabilities are 'socially generated,' they are equally 'open to deliberate change through reform and civic engagement'. By treating resilience as a proactive project rather than a passive by-product of wealth, these cases prove that democracies can effectively 'reclaim epistemic agency' and transform potential sites of vulnerability into bastions of resistance.

Table 2: Democratic Vulnerability to Cognitive Warfare.

Note. Composite scores were calculated by the authors based on the methodology described in the Research Design section. Component scores are derived from data from Reporters Without Borders (2025), the Reuters Institute for the Study of Journalism (2025), Transparency International (2025), the World Bank Worldwide Governance Indicators (2025), the Varieties of Democracy (V-Dem) Institute (2025), and OECD educational statistics (2025). Each case is scored on three dimensions using a standardised 0–3 scale, where 0 indicates low structural vulnerability and 3 indicates high structural vulnerability. Epistemic Infrastructure Decay (EID) captures the erosion of institutions that support shared knowledge production (example, media concentration, collapse of local journalism, platform‑dominated news). Institutional Capture (IC) reflects the degree to which regulatory bodies, public media, and oversight institutions are politicised or subject to revolving‑door influence. Democratic Agency Erosion (DAE) measures the weakening of citizens’ capacity for critical engagement, including economic precarity, defunded civic and humanities education, and exposure to digital manipulation. Cultural‑Identity Fragmentation (CIF) and Crisis Governance Patterns (CGP) act as contextual modifiers, increasing scores where polarised identity politics and crisis‑driven rule create additional opportunities for cognitive operations.

The empirical findings presented in Table 2 provide quantitative support for this paper's central thesis. The data reveals a clear clustering of nations that aligns with our theoretical framework. The designated negative cases, Finland (Composite Score: 1.30) and Switzerland (1.67), anchor the low end of the spectrum, empirically confirming their high degree of structural resilience. In stark contrast, the primary case studies—Bulgaria (7.74), Hungary (7.51), Brazil (6.00), and the United States (4.50)—exhibit moderate to severe vulnerability, reflecting deep-seated decay across their epistemic, institutional, and civic infrastructures. In the Philippines, the high vulnerability score reflects documented media capture and institutional weakening under the Duterte and Marcos administrations, including legal and regulatory pressure on independent outlets. Combined with persistent corruption, politicized oversight, and a highly social-media-dependent information environment, this supports elevated EID, IC, and DAE scores consistent with the broader high-risk cluster. This evidence suggests that a democracy's susceptibility to cognitive warfare is not arbitrary but is systematically correlated with its internal structural integrity, confirming that endogenous weaknesses are the critical determinants of democratic resilience. Crucially, these vulnerabilities are not simply inherent to democracy but are enabled and amplified by decades of political and economic choices. While often linked to neoliberal reforms, our comparative analysis—spanning Western, Central/Eastern European, and non-Western cases—suggests that CW exploits a broader set of structural weaknesses. Populist governance, persistent inequality, and weak regulatory

traditions, as seen in India, Brazil, and South Africa, also create openings for manipulation. This research advances the field by reframing CW as a systemic challenge rather than a purely technological or external threat. Our findings reinforce the argument that technical countermeasures are necessary but insufficient; without addressing the underlying structural drivers of epistemic collapse, institutional decay, and agency erosion, democracies will remain fundamentally susceptible (Giroux 2014).

This study's primary contribution is a diagnostic toolkit for identifying and addressing CW vulnerabilities, prioritizing structural interventions over reactive measures. However, any policy response must navigate normative tensions. Efforts to increase algorithmic transparency and regulate content risk enabling censorship, as seen in India’s IT Act 2021 and Singapore’s Protection from Online Falsehoods and Manipulation (POFMA). Effective safeguards must therefore include sunset clauses, independent multi-stakeholder oversight, and a narrow focus on procedural transparency to protect free expression while mitigating manipulation.

Recommendations

To address the multi-level vulnerabilities identified, we propose the following targeted recommendations:

Rebuild Epistemic Infrastructure:

Mandate algorithmic transparency through public registries and independent audits for major platforms.

Reinvest in public-interest journalism and independent media, funded by digital levies on large tech companies.

Expand support for fact-checking organizations and integrate media literacy into school curricula, drawing on successful models from Finland and Taiwan (OECD 2023).

Strengthen Institutional Safeguards:

Establish enforceable post-employment "cooling-off" periods for regulators to prevent capture.

Secure election infrastructure with independent security audits and open-source software mandates.

Foster multilateral counter-disinformation coalitions, such as the European Digital Media Observatory (EDMO), for intelligence sharing and rapid response.

Revitalize Civic Agency:

Implement digital hygiene and critical thinking curricula in secondary schools.

Adopt policies—such as labour protections and social safety nets—to reduce and acknowledge economic precarity.

Promote inclusive public spheres through participatory budgeting, community forums, and horizontal civil society networks (Hardt & Negri 2011) drawing inspiration from Porto Alegre and Lithuania’s “Elves.” (V-Dem Institute 2024)

Advance Research and Policy Integration:

Support interdisciplinary research on the evolving mechanisms of CW, following the model of NATO’s Cognitive Warfare Concept and Singapore’s Centre of Excellence for National Security (NATO Science and Technology Organisation 2023).

Integrate cognitive defence (psychological preparedness, strategic communications) into national security strategies, as implemented in Israel (Ya’ alon 2019).

Conclusions

This study set out to examine how internal structural decays in democratic societies—specifically within media, institutional, and civic infrastructures—create exploitable vulnerabilities for cognitive warfare (CW) operations. Our analysis demonstrates that these vulnerabilities are not an inherent flaw of democratic openness but are the result of cumulative structural transformations that have eroded democratic resilience. The evidence suggests that these vulnerabilities are not inevitable. Rather, they result from cumulative structural transformations, such as market-driven media landscapes and institutional hollowing, which have eroded democratic resilience. Our findings underscore that structural vulnerability to cognitive warfare is not solely a product of the amplification effect of neoliberalism, but can arise from a variety of political, economic, and social dynamics.

The multi-level framework advanced here illustrates that effective responses to cognitive warfare must be systemic and structural, not merely technical or reactive. Rebuilding resilience requires deliberate, simultaneous investment in epistemic infrastructure, robust institutional safeguards, and the renewal of civic agency. As this research has shown, democracies can reclaim epistemic agency through initiatives like public media reinvestment, algorithmic transparency mandates, and the fostering of horizontal civic networks, thereby transforming vulnerabilities into sites of resistance.

As cognitive warfare continues to evolve—leveraging AI, big data, and psychological operations—the stakes for democratic societies will only grow. The ultimate challenge is not simply to defend against manipulation, but to restore the conditions for collective self-governance: shared truth, institutional trust, and meaningful civic participation. Only by

addressing the full spectrum of these structural drivers can democracies restore resilience and reclaim openness as a strength.

Bibliography

Abernathy, PM 2020, ‘The expanding news deserts.’ Centre for Innovation and Sustainability in Local Media, UNC Hussman School of Journalism and Media. viewed 4 March 2026, https://www.usnewsdeserts.com/reports/news-deserts-and-ghost-newspapers-will-local-news-survive/.

Acemoglu, D & Robinson, JA 2019, The narrow corridor: States, societies, and the fate of liberty. New York: Penguin Press. viewed 4 March 2026, https://samataindia.org.in/data/books/Daron-Acemogulu-James-A-Robinson_The-Narrow-Corridor.pdf.

Adler, E 1997, ‘Seizing the middle ground: Constructivism in world politics.’ European Journal of International Relations, 3(3), 319–363. viewed 4 March 2026, https://doi.org/10.1177/1354066197003003003.

Adler-Nissen, R & Pouliot, V 2026, ‘Power in practice: Negotiating the international intervention in Libya.’ European Journal of International Relations, 20(4), 889–911. viewed 4 March 2026, https://doi.org/10.1177/1354066114543209.

Allen, K & le Roux, J 2024, Under the influence? Online mis/disinformation in South Africa’s May 2024 election. Institute for Security Studies (ISS) Africa. viewed 4 March 2026, https://issafrica.org/research/southern-africa-report/under-the-influence-online-mis-disinformation-in-south-africa-s-may-2024-election.

Arendt, H 1967, Truth and politics. In Between past and future: Eight exercises in political thought. Viking Press. viewed 4 March 2026, https://german.yale.edu/sites/default/files/arendt.truth_and_politicslying_in_politics.pdf.

Arquilla, J & Ronfeldt, D 1996, The advent of netwar. RAND Corporation. viewed 4 March 2026, https://www.rand.org/pubs/monograph_reports/MR789.html.

Atlantic Council. 2018, ‘Defining Russian election interference: An analysis of select 2014 to 2018 cyber enabled incidents’. viewed 4 March 2026, https://www.atlanticcouncil.org/in-depth-research-reports/issue-brief/defining-russian-election-interference-an-analysis-of-select-2014-to-2018-cyber-enabled-incidents-2/.

Avritzer, L 2017, ‘Participation in democratic Brazil: From popular hegemony and innovation to middle-class protest.’ Opin. Pública, 23(1). viewed 4 March 2026, https://doi.org/10.1590/1807-0191201723143.

Bajomi-Lázár, P 2024, ‘Media capture theory: A paradigm shift?’, Central European Journal of Communication, 17(2(36)), 238–246. viewed 4 March 2026,

https://doi.org/10.51480/1899-5101.17.2(36).567.

Belfer Centre for Science and International Affairs 2019, ‘Cognitive warfare: The Russian threat to election integrity in the Baltic States.’ viewed 4 March 2026,

https://www.belfercenter.org/publication/cognitive-warfare-russian-threat-election-integrity-baltic-states.

Benkler, Y Faris, R, & Roberts, H 2018, Network propaganda: Manipulation, disinformation, and radicalization in American politics. Oxford: Oxford University Press. https://doi.org/10.1093/oso/9780190923624.001.0001.

Berger, PL & Luckmann, T 1966, The social construction of reality: A treatise in the sociology of knowledge. New York: Anchor Books. viewed 4 March 2026, https://amstudugm.wordpress.com/wp-content/uploads/2011/04/social-construction-of-reality.pdf.

Bigo, D, Bonditti, P, & Olsson, C 2011, ‘Mapping the European field of security professionals.’ In D. Bigo, R. Bocco, & J.-L. Piermay (Eds.), Europe’s 21st century challenge: Delivering liberty (pp. 15–36). London: Routledge. viewed 4 March 2026, https://doi.org/10.4324/9781315581095-4.

Blyth, M 2002, Great transformations: Economic ideas and institutional change in the twentieth century. Cambridge: Cambridge University Press. viewed 4 March 2026, https://doi.org/10.1017/CBO9781139087230.

Bradshaw, S, & Howard, PN 2019, ‘The global disinformation order: 2019 global inventory of organised social media manipulation.’ Computational Propaganda Research Project, Oxford: University of Oxford. viewed 4 March 2026, https://demtech.oii.ox.ac.uk/wp-content/uploads/sites/12/2019/09/CyberTroop-Report19.pdf.

Brooks, R 2021, The real scandal? The revolving door between government and business is still open. The Guardian. April 15. viewed 4 March 2026, https://www.theguardian.com/commentisfree/2021/apr/15/real-scandal-revolving-door-between-government-and-business-still-open-.

Brown, W 2015, Undoing the demos: Neoliberalism's stealth revolution. New York: Zone Books. viewed 4 March 2026, https://doi.org/10.2307/j.ctt17kk9p8.

Canadian Centre for Cyber Security 2023, National Cyber Threat Assessment 2023–2024. viewed 4 March 2026, https://www.cyber.gc.ca/en/guidance/national-cyber-threat-assessment-2023-2024.

Center for the Study of Democracy 2025, Seizing the edge in cognitive warfare. viewed 4 March 2026, https://csd.eu/blog/blogpost/2025/07/03/seizing-the-edge-in-cognitive-warfare/.

Centre for Media Pluralism and Media Freedom (CMPF) 2016, Media Pluralism Monitor 2016. European University Institute. https://digital-strategy.ec.europa.eu/en/library/media-pluralism-monitor-2016-results.

Centre for Media Pluralism and Media Freedom 2024, Media Pluralism Monitor 2024: Bulgaria Country Report. European University Institute. viewed 4 March 2026, https://cmpf.eui.eu/country/bulgaria/

Couldry, N & Mejias, UA 2019, The costs of connection: How data is colonizing human life and appropriating it for capitalism. Stanford, CT: Stanford University Press. https://www.sup.org/books/sociology/costs-connection.

Crouch, C 2004, Post-democracy. Cambridge: Polity Press. https://www.politybooks.com/bookdetail?book_slug=post-democracy--9780745633145.

d’Ancona, M 2017, Post-Truth: The New War on Truth and How to Fight Back. London: Ebury. https://archive.org/details/posttruthnewwaro0000danc.

Deppe, C & Schaal, GS 2024, ‘Cognitive warfare: A conceptual analysis of the NATO ACT model.’ Frontiers in Big Data, 7, Article 1452129. viewed 4 March 2026, https://doi.org/10.3389/fdata.2024.1452129.

DiResta, R et al. 2021, ‘The long fuse: Misinformation and the 2020 election.’ Stanford Internet Observatory. viewed 4 March 2026, https://fsi.stanford.edu/publication/long-fuse-misinformation-and-2020-election.

EU DisinfoLab. 2023, Disinformation landscape in Romania. viewed 4 March 2026, https://www.disinfo.eu/wp-content/uploads/2023/09/20230919_RO_DisinfoFS.pdf.

European Audiovisual Observatory 2019, ‘Independence of media regulators. Council of Europe. viewed 4 March 2026, https://www.obs.coe.int/en/web/observatoire/publications/-

/asset_publisher/BEt0nsJbKkpD/content/the-independence-of-media-regulatory-authorities-in-europe.

European External Action Service 2022, ‘First report on foreign information manipulation and interference threats.’ viewed 4 March 2026, https://www.eeas.europa.eu/eeas/first-report-foreign-information-manipulation-and-interference-threats_en.

Evans, SD et al. 2023, ‘From empowerment to community power in participatory budgeting.’ American Behavioural Scientist. 67(4), pp.578-592. viewed 4 March 2026, https://journals.sagepub.com/doi/abs/10.1177/00027642221086947.

Finnemore, M & Sikkink, K 1998, ‘International norm dynamics and political change.’ International Organization, 52(4), pp. 887–917. viewed 4 March 2026, https://www.jstor.org/stable/2601361.

Freedom House 2024, ‘Freedom in the World 2024: The mounting damage of flawed elections and armed conflict.’ viewed 4 March 2026, https://freedomhouse.org/report/freedom-world/2024/mounting-damage-flawed-elections-and-armed-conflict.

Gandini, A, Keeling, S, & Reviglio, U 2025, ‘Conceptualising the ‘algorithmic public opinion’: Public opinion formation in the digital age.’ Dialogues on Digital Society, 0(0). viewed 4 March 2026, https://doi.org/10.1177/29768640251323147.

Garz, M, & Ots, M 2025, ‘Media consolidation and news content quality.’ Journal of Communication, 75(3), pp. 195-220. viewed 4 March 2026, https://doi.org/10.1093/joc/jqae053.

Giroux, HA 2014, Neoliberalism's war on higher education. London: Haymarket Books. viewed 4 March 2026, https://dialnet.unirioja.es/descarga/articulo/6612122.pdf.

GLOBSEC 2023, ‘GLOBSEC Trends 2023: Central and Eastern Europe one year into the war.’ viewed 4 March 2026, https://www.globsec.org/what-we-do/publications/globsec-trends-2023.

Habermas, J 1984, ‘The theory of communicative action.’ Vol. 1: Reason and the rationalization of society (T. McCarthy, Trans.). New York: Beacon Press. (Original work published 1981). viewed 4 March 2026, https://teddykw2.wordpress.com/wp-content/uploads/2012/07/jurgen-habermas-theory-of-communicative-action-volume-1.pdf.

Haigh, KZ & Andrusenko, J 2021, Cognitive electronic warfare: An artificial intelligence approach. Norwood, MA: Artech House. ISBN: 978-1-63081-811-1. viewed 4 March 2026, https://ieeexplore.ieee.org/document/9538834/.

Hardt, M & Negri, A 2011, Commonwealth. Harvard, MA: Harvard University Press. viewed 4 March 2026, https://selforganizedseminar.wordpress.com/wp-content/uploads/2011/08/hardt_negri_commonwealth.pdf.

Hopf, T 2013, ‘Common-sense constructivism and hegemony in world politics.’ International Organization, 67(2), pp. 317–354. viewed 4 March 2026, https://doi.org/10.1017/S0020818313000040.

Hopkins, AM, Menz, BD, & Sorich, MJ (2024). Potential of Large Language Models as Tools Against Medical Disinformation-Reply. JAMA internal medicine, 184(4), 450–451. https://doi.org/10.1001/jamainternmed.2024.0023.

Human Rights Watch 2022, ‘Events of 2021. In World Report 2022.’ viewed 4 March 2026, https://www.hrw.org/world-report/2022/country-chapters/Philippines.

Human Rights Watch 2025, ‘From bad to worse: The deterioration of media freedom in Greece.’ viewed 4 March 2026, https://www.hrw.org/report/2025/05/08/bad-worse/deterioration-media-freedom-greece.

Ibrahim, F, Rhode, S, & Daseking, M 2023, ‘A systematic review of cognitive and psychological warfare.’ The Driehaus HRO Journal. viewed 4 March 2026, https://tdhj.org/blog/post/cognitive-psychological-warfare/.

International Press Institute (IPI) 2022, ‘Media freedom in Hungary ahead of 2022 election: A mission report.’ viewed 4 March 2026, https://ipi.media/mission-report-media-freedom-in-hungary-ahead-of-2022-election/.

Iyengar, S, Lelkes, Y, Levendusky, M, Malhotra, N, & Westwood, SJ 2019, ‘The origins and consequences of affective polarization in the United States.’ Annual Review of Political Science, 22, pp. 129–146. viewed 4 March 2026, https://doi.org/10.1146/annurev-polisci-

051117-073034.

Kahneman, D 2011, Thinking, fast and slow. New York: Farrar, Straus and Giroux. viewed 4 March 2026, https://mlsu.ac.in/econtents/2950_Daniel%20Kahneman%20-

%20Thinking,%20Fast%20and%20Slow%20(2013).pdf.

Krook, ML & True, J 2010, ‘Rethinking the life cycles of international norms: The United Nations and the global promotion of gender equality.’ European Journal of International Relations, 18(1), pp. 103-127. viewed 4 March 2026, https://doi.org/10.1177/1354066110380963.

Kuo, R & Marwick, A 2021, ‘Critical disinformation studies.’ Harvard Kennedy School Misinformation Review, 2(4). viewed 4 March 2026, https://doi.org/10.37016/mr-2020-76.

Laclau, E & Mouffe, C 1985, Hegemony and socialist strategy: Towards a radical democratic politics. London: Verso. viewed 4 March 2026, https://www.versobooks.com/products/1158-hegemony-and-socialist-strategy?srsltid=AfmBOoqJMrekl5wH0Sm_dngMnIWG419DS2mlTPPE_n_PzsFIOujTXiO  v.

Levitsky, S & Ziblatt, D 2018, How Democracies Die: What History reveals About Our Future. New York: Crown Publishing Group. viewed 4 March 2026, https://psi424.cankaya.edu.tr/uploads/files/Levitsky%20%26%20Ziblatt%2C%20How%20D  emocracies%20Die_%20What%20History%20Reveals%20about%20Our%20Future%20%2  82019%20--%20Penguin%20Books%29.pdf.

Lijphart, A 1971, Comparative politics and the comparative method. The American Political Science Review, 65(3), pp. 682–693. viewed 4 March 2026, https://doi.org/10.2307/1955513.

McIntyre, L 2018, Post-Truth. Cambridge, MA: MIT Press. viewed 4 March 2026, https://staibabussalamsula.ac.id/wp-content/uploads/2024/05/Post-Truth-staibabussalamsula.ac_.id_.pdf.

Marma, KJS 2025, ‘Cognitive warfare: The invisible frontline of global conflicts.’ Modern Diplomacy. viewed 4 March 2026, https://moderndiplomacy.eu/2025/02/12/cognitive-warfare-the-invisible-frontline-of-global-conflicts/.

Martin, GJ & McCrain, J 2019, ‘Local news and national politics.’ American Political Science Review, 113(2), pp. 372–384. viewed 4 March 2026, https://doi.org/10.1017/S0003055418000965.

McChesney, RW 2015, Rich media, poor democracy: Communication politics in dubious times (20th anniversary ed.). New York: The New Press. viewed 4 March 2026, https://experts.illinois.edu/en/publications/rich-media-poor-democracy-communication-politics-in-dubious-times/.

Miller, S 2023, ‘Cognitive warfare: An ethical analysis.’ Ethics and Information Technology, 25(3), p.35 viewed 4 March 2026, https://doi.org/10.1007/s10676-023-09717-7.

Miskimmon, A, O'Loughlin, B, & Roselle, L 2013, Strategic Narratives: Communication Power and the New World Order (1st ed.). London: Routledge. viewed 4 March 2026, https://doi.org/10.4324/9781315871264.

NATO Allied Command Transformation 2021, ‘Cognitive warfare: An attack on truth and thought. North Atlantic Treaty Organization.’ viewed 4 March 2026, https://www.act.nato.int/wp-content/uploads/2023/12/Cognitive-Warfare.pdf.

NATO Allied Command Transformation. 2023. Cognitive Warfare Concept: Exploratory Study. Norfolk, VA: NATO ACT. viewed 4 March 2026, https://www.act.nato.int/activities/cognitive-warfare/.

NATO Allied Command Transformation 2023, ‘Cognitive Warfare: NATO’s Sixth Operational Domain.’ viewed 4 March 2026, https://www.act.nato.int/activities/cognitive-warfare/.

NATO Science & Technology Organization 2022, ‘SAS-163 Research Symposium on Cognitive Warfare.’ viewed 4 March 2026,

https://www.nato.int/nato_static_fl2014/assets/pdf/2023/3/pdf/230327-NATO-STO-Highlights.pdf.

NATO Science & Technology Organization (NATO STO) 2023, ‘Cognitive Warfare Conceptual Series.’ viewed 4 March 2026, https://www.act.nato.int/activities/cognitive-warfare/.

Newman, N, Fletcher, R, Robertson, CT, & Nielsen, RK 2024, ‘Reuters Institute Digital News Report 2024.’ Reuters Institute for the Study of Journalism. viewed 4 March 2026, https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2024.

Novais, RAS da C, & Leite, ÂMT 2024, ‘Attack in the form of defence? Populist anti-media tactics to avoid being blamed.’ Palabra Clave, 27(4), e2747. viewed 4 March 2026, https://doi.org/10.5294/pacla.2024.27.4.7.

OECD 2020, ‘PISA 2018 Results (Volume VI): Are students ready to thrive in an interconnected world?’ OECD Publishing. viewed 4 March 2026, https://www.oecd.org/en/publications/pisa-2018-results-volume-vi_d5f68679-en.html.

OECD 2023, ‘Facts not fakes: Tackling disinformation, strengthening information ecosystems.’ OECD Publishing. viewed 4 March 2026, https://doi.org/10.1787/d9dae5b5-en.

OECD 2023, ‘Media literacy and resilience in Finnish schools.’ OECD Publishing. viewed 4 March 2026, https://www.oecd.org/en/publications/mis-and-disinformation_b00de6dc-en/media-literacy-education-system_d067f517-en.html.

OECD 2023, ‘The OECD framework for measuring and monitoring civic space.’ OECD Publishing. viewed 4 March 2026, https://doi.org/10.1787/1b241870-en.

Onuf, N 2012, World of our making: Rules and rule in social theory and international relations (1st ed.). London: Routledge. viewed 4 March 2026, https://doi.org/10.4324/9780203722428.

OpenAI 2024, ‘AI and covert influence operations: Latest trends.’ viewed 4 March 2026, https://downloads.ctfassets.net/kftzwdyauwt9/5IMxzTmUclSOAcWUXbkVrK/3cfab518e6b1  0789ab8843bcca18b633/Threat_Intel_Report.pdf.

Pariser, E 2011, The Filter Bubble. London: Penguin. viewed 4 March 2026, https://www.penguin.co.uk/books/181850/the-filter-bubble-by-pariser-eli/9780241954522.

Pamment, J & Bjola, C 2020, Countering online propaganda and extremism: The dark side of digital diplomacy. London: Routledge. viewed 4 March 2026, https://www.routledge.com/Countering-Online-Propaganda-and-Extremism-The-Dark-Side-of-Digital-Diplomacy/Bjola-Pamment/p/book/9781138578630.

Persily, N 2017, ‘Can democracy survive the Internet?’ Journal of Democracy, Vol 28, Issue 2, pp. 63-76. viewed 4 March 2026, https://law.stanford.edu/publications/can-democracy-survive-the-internet/.

Persily, N 2020, ‘The 2020 U.S. election: A perfect storm.’ Journal of Democracy, 31(2), pp.94–108. viewed 4 March 2026, https://doi.org/10.1353/jod.2020.0022.

Przeworski, A & Teune, H 1970, The logic of comparative social inquiry. Hoboken, NJ: Wiley-Interscience. viewed 4 March 2026, https://dn790000.ca.archive.org/0/items/in.ernet.dli.2015.138830/2015.138830.The-Logic-Of-Comparative-Social-Inquiry.pdf.

Reporters Without Borders. 2025, ‘2025 World Press Freedom Index.’ viewed 4 March 2026, https://rsf.org/en/index.

Reuters Institute for the Study of Journalism 2025, Digital News Report 2025. University of Oxford. https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2025.

Satterthwaite, D et al. 2018, ‘Porto Alegre: Participatory budgeting and the challenge of sustaining transformative change.’ World Resources Institute. viewed 4 March 2026, https://www.oidp.net/docs/repo/doc415.pdf.

Schmidt, VA 2008, ‘Discursive institutionalism: The explanatory power of ideas and discourse.’ Annual Review of Political Science, 11, pp.303–326. viewed 4 March 2026, https://doi.org/10.1146/annurev.polisci.11.060606.135342.

Seloane, T 2023, ‘Disinformation for hire: The new gig in Africa’s attention economy.’ Institute for Security Studies. viewed 4 March 2026, https://issafrica.org/research/policy-brief/disinformation-for-hire-the-new-gig-in-africas-attention-economy.

Shi, F et al. 2024, ‘Who is the most vulnerable to LLM-generated misinformation?’ arXiv. viewed 4 March 2026, https://arxiv.org/abs/2402.19423.

Snyder, T 2018, The road to unfreedom: Russia, Europe, America. Tim Duggan Books. viewed 4 March 2026, http://www.iwm.at/publication/books-by-fellows/the-road-to-unfreedom-russia-europe-america.

Sophos 2024, ‘Political manipulation with massive AI model-driven misinformation and microtargeting.’ Sophos News. viewed 4 March 2026, https://news.sophos.com/en-us/2024/10/02/political-manipulation-with-massive-ai-model-driven-misinformation-and-microtargeting/.

Splidsboel Hansen, F 2021, Russian hybrid warfare: A study of disinformation. DIIS Report 2021:09, Danish Institute for International Studies. viewed 4 March 2026, https://www.econstor.eu/bitstream/10419/197644/1/896622703.pdf.

Standing, G 2011, The precariat: The new dangerous class. London: Bloomsbury Academic. viewed 4 March 2026, https://library.oapen.org/bitstream/handle/20.500.12657/58691/9781849664547.pdf?sequenc  e=1&isAllowed=y.

Stigler, GJ 1971, ‘The theory of economic regulation.’ The Bell Journal of Economics and Management Science, 2(1), pp. 3–21. viewed 4 March 2026, https://doi.org/10.2307/3003160.

Sunstein, CR 2017, #Republic: Divided democracy in the age of social media. Princeton, NJ: Princeton University Press. viewed 4 March 2026, https://press.princeton.edu/books/hardcover/9780691175515/republic?srsltid=AfmBOoq4PJb  emtOTI0W9sY8kgJ1uFfN33NLtIlKiSGwJE1trg19rmia4.

TekSavvy Solutions Inc. 2023, ‘Part 1 Application by TekSavvy Solutions Inc. Concerning Undue Preference.’ Canadian Radio-television and Telecommunications Commission (CRTC). viewed 4 March 2026, https://www.teksavvy.com/wp-content/uploads/2023/01/TekSavvy-Part-1-Application-Concerning-Undue-Preference-January-19.pdf.

The Heritage Foundation 2024, ‘2024 Index of Economic Freedom.’ viewed 4 March 2026 https://www.heritage.org/index/.

The Revolving Door Project 2024, ‘Unmasking FCC's revolving door with telecom giants.’ viewed 4 March 2026, https://therevolvingdoorproject.org/unmasking-fccs-revolving-door-with-telecom-giants/.

Transparency International EU 2020, ‘Access all areas: When EU politicians become lobbyists.’ viewed 4 March 2026, https://transparency.eu/access-all-areas/.

Transparency International 2024, ‘Corruption perceptions index 2023.’ viewed 4 March 2026 https://www.transparency.org/en/cpi/2023.

Transparency International 2025, ‘Corruption perceptions index 2024.’ viewed 4 March 2026 https://www.transparency.org/en/cpi/2024.

Varieties of Democracy (V-Dem) Institute 2025, ‘V-Dem Dataset v15.’ University of Gothenburg. viewed 4 March 2026, https://www.v-dem.net/data/the-v-dem-dataset/.

V-Dem Institute 2024, ‘Democracy report 2024: Democracy winning and losing at the ballot.’ University of Gothenburg. viewed 4 March 2026, https://v-dem.net/documents/43/v-dem_dr2024_lowres.pdf.

Vykopal, I, Pikuliak, M, Srba, I, Moro, R, Macko, D, & Bielikova, M 2024, ‘Disinformation capabilities of large language models.’ In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Association for Computational Linguistics. viewed 4 March 2026, https://doi.org/10.18653/v1/2024.acl-long.793.

Welch, DA 2013, Painful choices: A theory of foreign policy change. Princeton, NJ: Princeton University Press. viewed 4 March 2026, https://press.princeton.edu/books/hardcover/9780691123400/painful-choices?srsltid=AfmBOooKXYzypZdxdc5kT3TKmYfj6-Rw7dbVguPZR7x-sYR_VjBA61jh.

Wendt, A 1992, ‘Anarchy is what states make of it: The social construction of power politics.’ International Organization, 46(2), pp. 391–425. viewed 4 March 2026, https://doi.org/10.1017/s0020818300027764.

Wendt, A 1996, ‘Identity and structural change in international politics.’ In Y. Lapid & F. Kratochwil (Eds.), The return of culture and identity in IR theory. Boulder: Lynne Rienner Publishers. viewed 4 March 2026, https://www.rienner.com/title/The_Return_of_Culture_and_Identity_in_IR_Theory.

Wight, C 2006, Agents, structures and international relations: Politics as ontology. Cambridge: Cambridge University Press. viewed 4 March 2026, https://doi.org/10.1017/CBO9780511491764.

Williams, P et al. 2024, ‘Disinformation Election (DisElect) 2024: A large-scale comparative study of large language models in election disinformation generation.’ arXiv. viewed 4 March 2026, https://arxiv.org/abs/2403.04403.

World Bank Group 2025, ‘Worldwide Governance Indicators.’ viewed 4 March 2026, https://info.worldbank.org/governance/wgi/.

World Bank 2024, ‘Worldwide governance indicators.’ viewed 4 March 2026, https://info.worldbank.org/governance/wgi/

World Values Survey 2023, ‘WVS Wave 7 (2017-2022)’ viewed 4 March 2026, https://www.worldvaluessurvey.org/WVSContents.jsp.

Wright, CS 2025, ‘Cognitive castes: Artificial intelligence, epistemic stratification, and the dissolution of democratic discourse.’ arXiv. viewed 4 March 2026, https://arxiv.org/html/2507.14218v1.

Ya'alon, M 2019, The cognitive war as an element of national security: Based on personal experience. Institute for National Security Studies (INSS). viewed 4 March 2026, https://www.inss.org.il/wp-content/uploads/2019/10/yaalon.pdf.

Yazaki, Y 2019, ‘Economic and political impacts of newspaper mergers: Evidence from Japan.’ Social Science Research Network (SSRN). viewed 4 March 2026, https://doi.org/10.2139/ssrn.4072358.

Zgut-Przybylska, E 2024, ‘Pro-Kremlin disinformation is distorting democracy in Central and Eastern Europe.’ Review of Democracy. viewed 4 March 2026, https://revdem.ceu.edu/2024/10/08/pro-kremlin-disinformation-cee/.

Zhirnov, A et al. 2024, ‘Precarity and populism.’ European Sociological Review, 40(4), pp. 704–720. viewed 4 March 2026, https://doi.org/10.1093/esr/jcad052.

Zuboff, S 2019, ‘The age of surveillance capitalism: The fight for a human future at the new frontier of power.’ PublicAffairs. viewed 4 March 2026, https://www.hbs.edu/faculty/Pages/item.aspx?num=56791.