Photo Source: National Cyber Security Authority
By Julie Boury
When President Macron described a digital “wild west” for children at the European Digital Sovereignty Summit, socialization and historical knowledge rushed in to depict restrictive and urgent policies as common sense (TRT, 2025). In Power: A Radical View, Lukes (2005) challenges pluralist and neo-elitist conceptions of power by going beyond visible forms of conflict to examine “the existence of power as the imposition of internal constraints” (p. 13). But how does Lukes’ three-dimensional view of power fare when applied to our understanding of youth online safety policy? In this article, I will argue that, despite inherent methodological and practical limitations for a multi-level policy challenge, Lukes’ concept provides a unique analytical lens to identify how narratives become ‘sites of political maneuvering’ in youth online safety policy, providing an evolutive model for the construction of children’s online governance as well as the interrogation of the epistemological foundations of the definition of power (Zaman et al., 2020). After outlining the theoretical and empirical scope, I will first illustrate how online platforms shift safety discourse towards parents’ responsibility before arguing that states co-opt ‘moral panic’ and emotional discourse to encourage the selective surfacing of grievances deemed worthy of regulation. Finally, I will consider the methodological limitations of Lukes’ theory in quantifying and comparing transnational and localized preference-shaping power and defend the evolutive and interrogative value of his analytical framework.
This article concerns itself with power as preference-shaping, developed by Lukes as the complementary ‘third’ dimension to pluralist understandings of power as decision-making and neo-elitist views of power as agenda-setting. A crucial underlying assumption of all three dimensions is the existence of a counterfactual world in which actors would have chosen differently, if not for the power exercised. Also central to Lukes’ conception of preference-shaping power is the differentiation between an individual’s real interests and their preferences. Youth online safety policy further complicates this already contentious binary by distancing voters and adult participants from the real interests of the children whose behavior it seeks to regulate, opening up the possibility of Davison’s (1983) ‘third-person effect’ in communication whereby people believe the media to have a larger impact on others than themselves. Emerging in the 2000s as an effort to criminalize specific harms like sexual misconduct, youth online safety policy has since shifted with the rise of social media platforms to focus on protective measures for children’s data privacy online. Debates today are increasingly centered around restrictive measures, including age verification and blanket bans on social media use.
Confronted with the challenge youth online safety policy, Lukes’ (2005) three-dimensional view of power seeks to understand through which mechanisms the “willing compliance” of children and adults to domination by online platforms over the former’s data, behaviors, and ‘safety’ is secured (p. 10). Utilizing two frames, I will argue that youth online safety policy illustrates what Lukes (2005) describes as a latent conflict in which discursive and narrative power construct a particular understanding of children’s online ‘safety,’ thereby narrowing the scope of imaginable preferences. In turn, I examine how platforms shift online safety responsibilities towards parents and how state actors co-opt the emotional bearing of online harms to prioritize and sideline grievances. On the one hand, preference-shaping power is wielded by online platforms through narratives and ‘affordances’ that shift responsibility for children’s online safety to ‘digital parenting’. The depoliticized ‘affordances’ or features of online platforms enable the construction of a new understanding of children’s safety as the surveillance and systematic control by parents (Gibson, 2014; Fisk, 2016). For instance, in response to a series of lawsuits due to teens suicides, OpenAI introduced ‘parental controls’, including notifying parents when a child is distressed (OpenAI, 2025). Online harm to children is thus depicted as the result of specific harmful actors or inevitable ‘glitches’ rather than attributed to the platform and ICT infrastructures that enable and profit from these harms. Going beyond the second dimension of power which takes digital regulation off the agenda, this techno-determinism can only be explained by an effective discouragement or prevention “from having grievances” against online platforms altogether (Lukes, 2005, p. 11; Papageorgiou and Michaelides, 2016; Kania Lundholm, 2025). Lukes (2005) describes the mechanisms of this power as the “processes of socialization,” understood here as the web’s promises of opportunity and mass adoption vehiculated online and offline (p. 27). This power relies on value-laden narratives of ‘digital parenting’, exemplified by Australian parents’ expression of guilt for struggling to protect their children from online harms (Zaman et al., 2020). Similarly to how oil and gas companies shift climate responsibilities onto individuals through individual carbon footprint calculators, Lukes’ third dimension draws out how online platforms use depoliticized technological affordances and narratives of ‘digital parenting’ to shift responsibility for the protection of children online towards families.
On the other hand, power is wielded by state actors, especially those outside of the United States and China, by co-opting the ‘moral panic’ caused by online harms to advance blanket restrictions in displays of strength unaligned with youth’s real interests. Lukes’ lens helps understand the role played by state-led narratives in the unequal visibility afforded to different grievances. Indeed, in response to the rising power of US and Chinese-owned online platforms in shaping public discourse and economic flows, countries such as the United States, France, Australia, and the UK have displayed an appropriation of the emotional potency of online harms to children to support restrictive kinds of blanket policies. As Fisk (2016) puts it, “no one can say no” to the protection of children, as children are viewed as a particular subset of the population that must be protected and held under a constant supervisory gaze (Foucault, 1977; Hay, 2002). In addition, Zaman et al. (2020) demonstrate how coverage of high-profile cases in legacy media, such as suicides or ‘grooming,’ tends to mobilize fear and protective instincts. State actors wield this emotional power through bellicose and possessive vocabulary, seen in Gabriel Attal’s (2025) reminiscence of COVID-19 emergency measures in his ‘state of emergency’ and digital ‘curfew’ policy for children (Renaissance, 2025). State actors thus exert power not by suppressing grievances altogether but rather by selecting which ones are deemed ‘acceptable,’ in this case encouraging more restrictive policies. Other grievances are obscured, such as potential privacy risks of age verification technologies or the unequal offline social environments making children more vulnerable to online harms (Fisk, 2016). In sum, these two angles help understand the usefulness of Lukes’ third dimension of power in identifying the latent conflicts between state actors, online platforms, and children’s real interests that shape the perception of ‘acceptable’ or desirable policy for children’s online safety. It helps explain both the reticence to enact any concrete forms of platform regulation, with technology perceived as inevitable and even beneficial, and the rising popularity of more restrictive policies that risk exposing children to further surveillance (Wisniewski et al., 2022).
Having established the third dimension’s usefulness in uncovering the shaping, invisible force of these narratives on public policy preferences, I argue that Lukes’ three-dimensional view lacks the methodological tools to paint a complete picture of youth online safety policy while acknowledging that its evolutive framework provides valuable guidance to policy measures while contributing to an contestation and expansion of analytical tools. First, Lukes’ “remarkable disdain for methodological considerations” (Hay, 2002, p. 170) makes it difficult to examine the simultaneously globalized and culturally-anchored nature of youth online safety policy, which operates with multi-level governance frameworks spanning the UN Convention on the Rights of the Child (1989), the EU’s BIK+ strategy and national or regional initiatives (European Commission, 2022). Zaman et al. (2020) have demonstrated how child online safety narratives are particularly grounded in time and space, varying significantly for instance between the Australian and Belgian context. Following Fisk (2016), I argue that Lukes’ macro level perspective risks obscuring how micro-interactions might show more tactical resistance to children’s online harm narratives. In addition, although quantitative social science methods, including ones mobilizing LLMs, offer new opportunities for the massification of discourse-analysis, Lukes’ lens offers no methodology for quantifying these preference-shaping powers exercised by online platforms or states over children, limiting its empirical applications (Zaman et al., 2020; Argyle et al., 2023; Ingber and Su, 2025). Additional limitations lie beyond the scope of this article, including the normatively tenuous task of determining and distinguishing the ‘real interests’ of children from highly varied backgrounds and the assumption that power can never be exercised responsibly. Despite these considerable empirical and conceptual limitations, I argue that the value of Lukes’ (2005) third dimension of power holds because it offers an evolutive and “essentially contested” framework of critique that can be applied to both the policy challenge and the discipline itself (p. 14). Indeed, it implies the need for youth online safety policy to center the real interests of those it serves, for instance by including children in the governance process, and for a perpetual questioning of the power relations of proposed legislation. Turning towards the discipline, the value of Lukes’ view equally lies in its interrogation of who benefits from narrow, visible, and quantifiable definitions of power, pioneering an alternative and inter-disciplinary analysis. Indeed, Lukes breaks down an artificial separation between strictly behavioral analysis and more critical approaches to the lived experiences and linguistics of power which serves to further occult forces operating through more diffuse and invisible forms of influence on policy. As such, Lukes’ lens remains the historical harbinger of an intellectual turn in analytical approaches to the exercise of power and an imperfect but useful tool for policy and definitional contestation.
Overall, this article has argued that Lukes’ view of power as preference-shaping effectively highlights the discursive and productive power wielded by both online platforms and states in shaping the range of ‘acceptable’ policy preferences for children’s online safety. The lens goes beyond decision-making and agenda-setting to understand who benefits from harmful online content and extractive logics affecting both children and adults, and from certain strongman policy decisions. Though it fails to quantify these multi-level flows of power or suggest a legitimate exercise of discursive power, Lukes’ acknowledgement of the ever-evolving and incomplete nature of his theory challenges broader epistemological underpinnings of definitional approaches to power. As such, the third dimension offers not a definitive analytical framework, but rather one for continuing the contestation of power relations, whether in a policy challenge or within the analytical field.
Works Cited
Argyle, L.P. et al. (2023) “Leveraging AI for democratic discourse: Chat interventions can improve online political conversations at scale,” Proceedings of the National Academy of Sciences, 120(41), p. e2311627120. Available at: https://doi.org/10.1073/pnas.2311627120.
Attal, G. (2025) Gabriel Attal et Marcel Rufo : «Il faut interdire l’accès des jeunes de moins de 15 ans aux réseaux sociaux», Le Figaro. Available at: https://www.lefigaro.fr/vox/societe/gabriel-attal-et-marcel-rufo-il-faut-interdire-l-acces-des-jeunes-de-moins-de-15-ans-aux-reseaux-sociaux-20250429 (Accessed: November 28, 2025).
Davison, P. (1983) “The Third-Person Effect in Communication,” Public Opinion Quarterly, 47(1), pp. 1–15. Available at: https://doi.org/10.1086/268763.
European Commission (2022) European strategy for a better internet for kids – BIK+ | Shaping Europe’s digital future, Digital-strategy.ec.europa.eu. Available at: https://digital-strategy.ec.europa.eu/en/policies/strategy-better-internet-kids (Accessed: November 27, 2025).
Fisk, N.W. (2016) Framing Internet Safety: The Governance of Youth Online. The MIT Press. Available at: https://doi.org/10.7551/mitpress/10311.001.0001.
Foucault, M. (1977) Discipline and Punish: The Birth of the Prison. Vintage Books.
Gibson, J.J. (2014) The Ecological Approach to Visual Perception: Classic Edition. New York: Psychology Press. Available at: https://doi.org/10.4324/9781315740218.
Ingber, A.S. and Su, C.C. (2025) “Protecting Children of the TikTok Era: A Discourse Analysis in the Absence of Law,” Policy & Internet, 17(2), p. e431. Available at: https://doi.org/10.1002/poi3.431.
Introducing parental controls (2025). Available at: https://openai.com/index/introducing-parental-controls/ (Accessed: November 27, 2025).
Kania Lundholm, M. (2025) “Looking beyond the Glitch: What can we Learn from Sociologically Informed Critical Studies of Datafication?,” Sociologisk Forskning, 62(1–2), pp. 169–186. Available at: https://doi.org/10.37062/sf.62.27832.
Lukes, S. (2005) Power: A radical view. 2nd ed. Basingstoke: Palgrave Macmillan.
Papageorgiou, T. and Michaelides, P.G. (2016) “Joseph Schumpeter and Thorstein Veblen on technological determinism, individualism and institutions,” The European Journal of the History of Economic Thought, 23(1), pp. 1–30. Available at: https://doi.org/10.1080/09672567.2013.792378.
Renaissance (2025) Régulation des écrans et des réseaux sociaux chez les adolescents, https://parti-renaissance.fr/. Available at: https://parti-renaissance.fr/petitions/regulation-ecrans-et-rs (Accessed: November 27, 2025).
TRT (2025) Macron calls the online world “Wild West” for European teenagers, urges measures. Available at: https://www.trtworld.com/article/ff90af219f1b (Accessed: November 28, 2025). Zaman, B. et al. (2020) “Opposing narratives about children’s digital media use: a critical discourse analysis of online public advice given to parents in Australia and Belgium,” Media International Australia, 176(1), pp. 120–137. Available at: https://doi.org/10.1177/1329878X20916950.
