Platform governance and civil society organisations: Tensions between reform and revolution continuum

I. Introduction

The long awaited EU Digital Services Act (DSA) came into force in November 2022 and fully applied across the whole of the European Union in February 2024. This regulation constituted the first systematic attempt to regulate platforms. We can think of the DSA as operating at two levels: at the first level it works through providing a set of mandatory but broad rules for different categories of platforms; and at the second level it provides for the development of a set of voluntary codes of conduct which contain specific guidance for the application of the regulations and rules. Both make provisions for the participation of civil society organisations (CSOs) though there is a lack of specific details as to how this will happen.

The development of voluntary codes is considered part of a broader shift from ‘command and control’ regulation towards the incorporation of informal and voluntary steering mechanisms for the operation of transnational corporations (Gorwa, 2019a). These mechanisms involve the participation of platforms themselves in forms of self regulation as well as the participation of CSOs representing diverse communities and interests. The inclusion of CSOs is rooted in the multistakeholder governance model pioneered by the World Summit on the Information Society (Raboy and Landry, 2005) that presents the critical evaluation by civil society of government information regulatory practices. The model of multistakeholder governance typically includes three kinds of stakeholders is part of the platform governance triangle model (Gorwa, 2019a;) which includes: states, NGOs including CSOs¹, and firms. While the platform governance triangle alludes to a multi-stakeholder model of governance, Gorwa points out that there are power asymmetries between stakeholders and that civil society plays a marginal role, often used to merely legitimise decisions (Cammaerts 200; Dutton, 2015; Gorwa, 2019).

In broad terms, the involvement of CSOs in multistakeholder governance has been steadily increasing since the 1990s, and in particular in areas that local or national governance is no longer able to comprehensively address, for example environmental regulation, migration, and human rights (Cammaerts, 2009; cf European Commission, 2001). As Cammaerts (2009) points out, establishing links with CSOs was spearheaded by the UN as a means for increasing transparency and accountability. This began as early as in 1945, when consulting civil society was embedded in Art 71 the 1945 Charter of the UN. This concept of civic dialogue was subsequently taken up by the European Commission, which, in a White Paper on European Governance, explicitly incorporated the involvement of civil society as part of a “reinforced culture of consultation and dialogue” (EC, 2001: 16, quoted in Cammaerts, 2009: 6). Since the DSA will incorporate more voluntary codes for platforms (Griffin and van der Maelen, 2023), the role of civil society and that of non governmental organisations is expected to be strengthened. However, to what extent do, or, indeed, can, they represent user interests and how precisely do they shape platform governance? In this context, this article focuses on a particular kind of actor within civil society, that of digital rights CSOs. Broadly speaking, digital rights CSOs are a subset of civil society that have tended to advocate for the protection of human rights impacted by the use of communications technologies including the internet and various media (Access Now, 2024). In the EU we have advocated in areas such as government and policing surveillance, including intelligence sharing and biometric policing; data governance, including processing without informed consent; and novel forms of rights harms attached to new technologies, including online harms. Issues have included internet shutdowns, digital identities, encryption, facial recognition technologies, recommender systems, and harmful contents. We note how all of these concerns can have a platformed dimension, in that platforms collect, retain, analyse, control and exchange data across all of these concerns. And so in particular we examine how digital rights CSOs are involved in platform governance, focusing primarily on their relationship with, and attitudes towards platforms, while also scrutinising more closely their approach to formulating policy proposals. Theoretically, this article seeks to contribute to platform governance, by positioning it within a neoliberal approach to governance, and by exploring unresolved tensions between different policy actors and their interests. Empirically, we pose the following overarching research question: how are digital rights CSOs positioned and participating in platform governance in Europe? We approach this question in an exploratory manner, drawing on in-depth interviews with key informants from five leading CSOs in digital rights and with extensive work on digital policy, supported by auto-ethnography and a document analysis of materials posted in the CSO websites.

Our argument proceeds as follows. First, we provide a brief overview of the model of multistakeholder governance as it emerged in the World Summit of the Information Society (WSIS) in the early 2000s. We then formulate our theoretical approach, which, drawing on Michel Foucault, positions civil society as part of the neoliberal approach to governance. We then proceed to our empirical analysis and discussion. Our findings suggest that a structuring dimension of digital rights CSOs operating in this space is an ideological division to which we refer as the ‘reform versus revolution’ continuum. While those closer to the ‘reform’ end are looking to make incremental changes to improve the way platforms operate, those closer to the ‘revolution’ end take a more radical view aiming to dissolve platforms and their business model. While this structuring division reflects positions that are critical in different ways, pragmatic issues around funding and the hegemonic role of platforms undermine CSOs’ ability to act.

II. Setting the context: multistakeholder governance and civil society

In this section, we discuss the involvement of civil society organisations in governance and we specify and justify our focus on digital rights CSOs. We begin by elaborating on the multi stakeholder governance model, and then focus more closely on the role of civil society. The discussion illustrates two key tensions in the involvement of civil society in processes of governance: the first tension stems from the power asymmetries between civil society and the other actors in multi stakeholder governance, namely states and private companies; and the second tension is linked to the great diversity of civil society organisations and their potentially conflicting priorities and interests.

While a multistakeholder model of governance was first introduced in the Earth Summit in Rio de Janeiro in 1992 (Backstrand and Kylsater, 2014), it was not until the UN-supported World Summit of the Information Society (WSIS) that it became more concrete. WSIS was a two-phase summit organised by the United Nations in 2003 and 2005, with the aim to establish a dynamic multi stakeholder framework, designed to build “a people-centred, inclusive, and development-oriented Information Society” (Unesco, undated). The WSIS model of multistakeholder governance aimed to bring governments, civil society, and the private sector together on an equal footing to address the challenges of governing the internet as a global integrated digital infrastructure. This was the first time that civil society was included as an active participant in addressing global challenges in information and communication (Raboy and Landry, 2005).

Despite the potential and promise of multistakeholder governance in the WSIS model, the reality was different. As Hintz (2007) notes, while WSIS promised full participation, the implementation fell short, with participants from the civil society often limited to observer roles and excluded from critical decision-making spaces. Although civil society working groups and thematic discussions allowed for some influence, their contributions were frequently sidelined as annexes rather than being incorporated into formal negotiations (Hintz, 2007). More broadly, significant challenges emerged, including persistent power imbalances, as governments resisted deeper integration of non-state actors, and structural barriers and divides that excluded smaller CSOs and grassroots groups due to funding and access limitations (Hintz, 2007; Cammaerts, 2009). While therefore CSOs are meant to participate on an equal footing, power imbalances often reduce their role to that of legitimating the process of governance (Costanza-Chock, 2003; Chakravartty, 2006).

In addition to the power differentials between CSOs, states and firms, there are significant power dynamics within civil society organisations. Civil society is not a monolithic entity but comprises a wide spectrum of organisations, including NGOs, grassroots movements, professional associations, trade unions, advocacy groups, and community-based networks. These represent diverse interests, priorities, and ideologies, ranging from human rights advocacy to developmental or technological concerns. Chakravartty (2006) showed that there was a clear North-South divide in WSIS, where the Northern CSOs were more focused on matters of identity and recognition, while CSOs from the global South/global majority countries faced barriers to participation and were more concerned with questions of resources and economic inequality. Focusing on power hierarchies within civil society, Chakravartty’s analysis emphasises the inadequacy of assuming that civil society inherently represents the public interest or marginalised voices. Rather, she considers civil society as an elite actor, whose agendas can be shaped by donor priorities, funding constraints, and institutional affiliations, which may not align with grassroots needs. In these terms, Chakravartty views civil society as both a site of advocacy and a space of contested power relations.

As part of these power relations, Costanza-Chock (2003) points to the ideological alliance of WSIS and its multistakeholderism with a neoliberal agenda. According to her, WSIS prioritised the privatisation of information and communication systems using intellectual property rights and more broadly consolidating the knowledge commons into proprietary systems owned by multinational corporations. State-firm partnerships were framed as solutions to bridge the digital divide, but they primarily facilitated corporate exploitation of underserved markets. In this context, civil society was conflated with the private sector, undermining grassroots representation and allowing corporations to dominate. However, while WSIS was seen as co-opted by interests of firms, digital rights civil society groups used it as a platform to advocate for communication rights and access and to challenge the neoliberal framework. For Costanza-Chock, WSIS was a battleground where neoliberal policies clashed with grassroots demands for inclusivity and equitable access.

More broadly, these analyses of WSIS illustrate that despite its intentions, multistakeholder governance is ridden with power inequalities and tensions between conflicting interests. Since digital rights civil society is also subjected to internal power hierarchies and diversity of views, its participation in governance is unlikely to yield any real difference in terms of values such as equality and justice for all. While this discussion made clear some of the historical precedents of digital rights CSOs in internet governance and the failure of this participation to live up to expectations, the dominance of platform firms saw new regulation that has a clear role for civil society.

III. The neoliberal turn’s implications for CSO interventions into platform governance

Legal and regulatory instruments at the national and European level are integral to platform governance in the EU. In particular, Gorwa (2019a; 2019b) understands platform governance as consisting of a set of multilayered relationships between key stakeholders, including platforms themselves, individual and business users, governments, and other social and political actors. The inclusion of these actors and instruments can be understood as part of a broader shift towards neoliberal governance. If the previous section focused on the events around WSIS, this section expands on the theoretical dimensions of the involvement of private actors, including corporations and civil society in governance. Based on Foucault (2008) we explore the (neo)liberal approach to governance, positing that the space created for CSOs emerged within this paradigm. In this sense, the involvement of stakeholders from civil society is linked to the delegation of state power into the hands of non-state actors and with shifts in political rationality and the exercise of power (Foucault, 2008). Theoretically, therefore, we locate the emergence of civil society as part of (neo)liberal governance and this comes with a set of structural tensions that can capture CSOs in ways that may end up reproducing and legitimating a system of governance that civil society was formulated to oppose or at least to mitigate its impact. In this sense, it is the structural position of civil society that is entangled with neoliberal governance, notwithstanding the ideological stance of some parts of civil society against neoliberal policies. The structural position occupied by civil society and the principles and values that underpin at least some part of it are in constant tension. In our discussion below, we pursue the argument that civil society as a sector is part of neoliberal governance and as such structurally impeded in its quest for a ‘people-centred’, democratic and equitable approach to governing technologies. This is especially clear in platform governance and in the regulatory approach of the European Union’s DSA.

The neoliberal model: fomenting civil society organisation’s integral role in (platform) governance

Although the neoliberal paradigm is well known (see Harvey, 2007) it is worth repeating its key characteristics: firstly, that state and government intervention is minimal but still present in order to safeguard both the rule of law, security of the population and the principles of free market; secondly, the creation of state-firm partnerships as a more efficient and effective way to organise. In this manner, neoliberalism introduces a new political rationality, that shifts the focus from regulating markets and individuals through law to governing through market mechanisms and self-regulation (Foucault, 2008).

Historically, in the liberal European context, civil society emerged out of the tensions between the state and the market and the division between the public and private (Colas, 2002). As the state divests its sovereign role, it creates a space for the organised part of civil society to uphold liberty and act as safeguard against state encroachment. In line with liberal views, the state should allow for different interest groups to formulate and pursue demands on behalf of their constituencies; these are often organised as civil society organisations (CSOs). As we moved from liberal to neoliberal positions, in this ‘marketplace of ideas’ the role of CSOs has shifted from upholding liberal moral values such as liberty and autonomy to one in which they compete for influence (Pyykkönen, 2015). While therefore in classic liberalism civil society emerged as an organised form of collective existence that regulates civic and political participation, within the neoliberal paradigm it becomes primarily a means by which it counterbalances and guards against state power from the perspective of different interest groups (Foucault, 2008). At the same time, as states in neoliberalism are more concerned with matters of security and with creating attractive conditions for capital, it falls upon civil society to advocate and pressure the state (and industry) for regulation to protect and guarantee citizen and more broadly human rights (Lipschutz, 2005). For Foucault (2008, pp. 2–13, 295–296), civil society emerged as part of responding to the questions of “how to govern,” “how to govern right” and “how not to govern too much” and it is inextricably linked to the (neo)liberal governmental rationality which continuously seeks to limit itself. CSOs are therefore another technology of government within neoliberalism, tasked with pushing against state encroachment while also promoting social cohesion and aiding those marginalised by conditions of intense competition. In doing so, however, they contribute to rather than challenge this system of government, because neoliberal governmentality would not be able to function without the support of civil society. Following Foucault, in the neoliberal model, the state is concerned with security and regulating for keeping markets open; the market and its logics are expanded across all parts of society; and civil society ends up having to deal with the fallout of the marketisation of everything, providing a facade of decentralisation and democratic governance even as systemic inequalities are perpetuated.

Returning to the platform multistakeholder governance model, while the location of CSOs in between states and platforms is taken to constitute an independent position outside the formal exercise of power, it may in fact be an integral part of it. CSOs operate by seeking to bring matters to the attention of the state and firms actors, where, in order to be addressed, they must be integrated into state and corporate rationalities (Lipschutz, 2005). In this sense, civil society operates as a legitimation mechanism not only because it adds a superficial layer of democratic participation, but also at a deep, structural level, as it effectively advocates for state and firms actors to deal with problems through incorporating them in their logics. Since in neoliberalism the state logic coincides with the market, this prevails. In other words, rather than questioning and limiting state and market power, civil society ends up reinforcing it. As Meiksins-Wood (1995) argues, civil society acts as a legitimation mechanism for (neo)liberal democracies by tackling some of the symptoms of inequality, while obscuring fundamental social divisions and systemic injustices. It is important to point out that this happens even if civil society organisations are ideologically opposed to neoliberalism.

This is not to say that civil society does not have a role to play in democratic politics but rather to highlight the structural tensions that circumscribe its operations. Theorists such as Walzer (1995; see also Gramsci, 1971) argue that CSOs can influence states and markets by wielding knowledge and norms as well as by shaping public opinion. Indeed, even Foucault (2019) recognized that counter conduct can and does emerge when social actors reflexively, creatively and occasionally radically disrupt, subvert or counter existing power structures. As Pyykkönen (2015, p. 23) put it, “actions in civil society always also entail a chance for change, for doing or thinking about things differently”. Following this line of thought (see also Lipschutz, 2005, p. 231) we theorise that CSOs can push against the expansion of the market as well as against state control and surveillance. But they are faced with a dilemma: they can act in ways that either unwittingly support these logics or deliberately oppose them. How and if this happens in the context of platform governance is one of the questions that animate this article.

CSOs’ unintended role in normalising platform activities

Terry Flew (2021) proposes a three-phase periodisation of platform governance: the early phase of unregulated libertarian internet; the phase of platformisation, characterised by the dominance of platforms; and a new period that we are entering currently, which sees the dawn of the regulated internet. In this phase, Flew (2021) argues, we observe that states are becoming increasingly important in regulating platforms. At the same time, platform governance involves a high degree of regulatory innovation and generally with a high level of civil society participation and public interest compared to other industries (Flew, 2021). As we discuss below, the quasi-formal inclusion of CSOs can be considered a regulatory innovation. On the other hand, the DSA still mainly upholds a self- or at most a co-regulatory model, with platforms seen as ‘mere conduits’ and offered protections. In this sense, while the DSA constitutes a binding form of regulation it may not necessarily reflect a power shift from platforms to regulatory bodies/states. It is worth highlighting Foucault’s (2007) insight on the division of labour between markets and states that reflects the liberal rationality of liberty of the markets and individuals and security of the population; in this sense, we are still within the same paradigm and the shift towards state involvement in platform regulation is a correction concerning the security of the population and concerning the proper functioning of the free market and especially taking action against monopolies (Popiel, 2024).

Focusing on the participation of CSOs in the governance of platforms, Celeste (2019) outlines their normative role as one of providing the basis for the creation of a digital constitution conferring fundamental digital rights. For Celeste (2019; Celeste, Heldt, and Keller, 2021), digital constitutionalism emerges in a context where platforms pose increasing risks for constitutional equilibria through challenging fundamental rights, and where increased system differentiation, such as for example, the differentiation between print/broadcast media and digital platforms, requires the development of a novel set of rules and norms. Similarly, Suzor (2018) argues that platform rules such as Terms of Service do not adequately articulate the rights of users and the responsibilities of platforms. He therefore proposes the use of the rule of law to set up the obligations of platforms and the limits of their governance power. This is what digital constitutionalism proposes to do and it is constructed as a collective task in which all citizens should engage. Indeed, Celeste and colleagues (2023) find that there is considerable convergence around a key set of principles, and moreover, that CSO proposals provide enough detail to develop a meaningful operationalisation.

In digital constitutionalism, the role of CSOs becomes less about pushing against encroachment and more about formulating in positive terms guidelines as to how platforms should operate. However, it is not clear how these positions proposed by CSOs can influence or shape state and platform policy, as they seem to constitute non-binding norms. Similarly, although most of these positions revolve around human rights standards, Celeste et al. (2023) point out that there are differing views as to how these can best be advanced. It is here that normative ideals meet the messy sociological reality. Further, the unintended consequence of CSO interventions into platform spaces appears to be how CSOs are inadvertently functioning to normalise the existence of platform business model monopolies by focusing on regulatory principles that may or may not be upheld. Nevertheless, the EU’s DSA seems to provide a broad framework for the participation of CSOs in platform governance even if it is still unclear how this might unfold in practice.

The DSA’s direct path to CSO governance participation

While for the most part the role of civil society in platform governance is to develop normative approaches (as in digital constitutionalism) or to advocate for rights and fundamental freedoms, for the first time the DSA provides a path for direct participation in governance rather than looking to exert influence and shape policy. In particular, its chief goal is to set clear rules to protect citizens and their fundamental rights while fostering “innovation, growth and competitiveness” (European Commission, n.d.). There are three areas where the DSA provides (but does not mandate) for civil society involvement: firstly, in efforts to monitor the enforcement of its rules (e.g. Recital 90); secondly, in co-designing mitigation measures and contributing to the drawing up of additional codes of conduct as and if required by the Commission (e.g. Article 45(2); and thirdly, in advising and providing expert support to the Commission (e.g. Recital 137).

The potential for direct involvement of CSOs in the DSA marks a departure from their typical role. On the one hand, this development recognises the importance of non-state and non-market actors in regulatory efforts. On the other hand, the quasi-formalised role allocated to CSOs raises important questions concerning their capacities, their independence and their structural position in platform governance.

More specifically, Article 22 of the DSA provides a strong contemporary example of this new role for digital rights CSOs and platform governance. Under the new regulation, national government Digital Services Coordinators can designate entities including CSOs as “trusted flaggers”. Indeed, CSOs tend to be the main applicants for trusted flagger status (Klotsonis, 2024). These trusted flaggers are responsible for detecting potentially illegal content and alerting the platforms. This potential insertion of CSOs directly into the governance machinery of the DSA reflects the neoliberal governance logic that allows states to delegate their power away to CSOs as non-state actors, thereby shifting the political rationality towards legitimation of co-regulatory systems, while not necessarily addressing the substance of the problems that regulation is called to resolve. Further, trusted flaggers are required to have demonstrated expertise, competence and, importantly, independence. However, there is no transparency criteria at national levels detailing how an entity's donor priorities, institutional affiliations, current relationships or other existing/historical channels with platforms will impact - or indeed support - their selection as trusted flaggers.²We have seen for example that certain CSO workers are invited to specific stakeholder events or are allocated specific roles because of the relationships that they have established with people in government and industry. These soft power relations in multi stakeholder environments might influence the selection of trusted flaggers and a lack of formal transparency around this recruitment process obscures understanding. In this sense, the double dependency of CSO organisations on both industry and government compromises their ability to act independently and turns them into an extension of the market and the state.

This discussion showed that in liberal and neoliberal societies, civil society is associated with the private sector and tasked with curbing the powers of the state, representing a technology of government within liberalism. As neoliberal policies became increasingly dominant, civil society became oriented towards managing the sharper edges of the market, taking over some of the state’s functions around citizen protection (Lipschutz, 2005). As a technology of government, civil society can end up legitimating the current status quo if it fails to question not only the power held and exercised by the state and the market but also its own role and involvement in power hierarchies. As CSOs appear more and more integrated in the EU regulatory landscape, it becomes important to understand their own perceptions of their role, and their approach to platforms and the policy making process.

In this context, this article focuses on CSOs concerned with the digital sphere, platforms and technologies, probing their role and in particular their contributions to and views on platform governance. In particular, the focus is on digital rights CSOs that follow the European Digital Rights (EDRi) agenda, that is, they aim to protect civil and human rights and to advance a vision of a world where technology serves people, society and the planet³. The research questions that animate this research are formulated as follows: how do CSOs contribute to platform governance? What are the structuring tensions, internal and external factors and parameters that delineate their operations and affect their practices? Can CSOs meaningfully contribute to reshaping platform governance or are they bound to act as a legitimation mechanism? If so, what are the conditions under which this meaningful contribution can happen? The following section discusses the research approach used to address these questions.

IV. Our research approach

The article employs a qualitative approach, making use of ethnographic methods, in depth interviews with key informants, together with document analysis. The chief focus of our research is on the role of digital rights CSOs in platform governance who broadly share a vision that coalesces around the challenges to protection against unlawful interference with human rights and fundamental freedoms that have materialised in the digital age, including the formation of very large platforms referred to collectively as Big Tech. These CSOs prioritise the promotion and protection of human rights in the digital age including particularly the right to privacy/personal data protection, free expression, association and peaceful assembly, and equality and non discrimination. Their mandates and programmes of work often take three forms including: (i) research and knowledge building about new technologies, including their human rights and social justice impacts and co-related regulatory requirements, (ii) advocating for regulatory solutions for new technologies in areas of law, policy and practice through parliamentary submissions, consultation engagement and litigation, (iii) outreach, campaigning, awareness raising and education, for the public generally but also as direct advocacy to the state and firms.

We have selected five of the most active digital rights CSOs in Europe, two of which operate globally, whose actions are likely to influence outcomes in platform governance. These CSOs advocate for democratic requirements attached to fundamental freedoms and social justice in an age of technology, including transparency, accountability, remediation and oversight - not just in terms of how platforms operate but also in how policy makers and regulators develop and implement policies. In this sense, these EU digital rights CSOs operate not only as instruments of accountability to government, per their historically constituted roles, but also to platforms themselves, whose regulatory innovations have outpaced government’s ability to respond effectively at speed and scale.

Autoethnography

Our empirical analysis is grounded in institutional ethnography as a qualitative research method (Smith, 1987). Unlike other research where researchers consider the working of groups that are external to them, our empirical approach is informed by insights obtained through a privileged insider perspective of one of the authors, who has a background of an EU and International CSO worker, employed as a Director by a digital rights CSO that shares the vision and aims outlined in the above section. She remains there today as a senior fellow. As such, she is therefore familiar with the processes, arguments, tactics deployed, and field-level impacts targeted for digital rights and regulations. Taber (2010) elaborates that Smith promotes autoethnographic methods to better elucidate the complex set of relations that are connected, organised and mediated via the privileged location of ruling relations - i.e. the ‘corporations, government bureaucracies, academic and professional discourses, mass media, and the complex of relations that interconnect them’ (Smith, 2005, p.11; Taber, 2010; Juris and Khasnabish, 2013). There are two ways this approach has been useful for the current study: (i) in conjunction with our theoretical perspective it provided an entry point and guided the questions asked; (ii) as with the document analysis, it provided validation and support for the issues raised by the informants. In this sense, our application of the autoethnographic insider view was not used as a data gathering method but rather to inform and validate the analysis of data collected through interviews and document analysis.

Our interviews with EU digital rights CSO workers

This work involves semi-structured interviews of five key informants, all of them senior workers from the five leading EU CSOs described above. Rather than looking to conduct a survey across a larger and representative sample of CSOs in this sphere, this article presents the findings of an initial small research exploration into how a handful of influential CSOs understand and reflect on their role and practices. To conduct the interviews, the authors were granted research ethics approval from their university on the basis of fully anonymising the informants and their organisation and obtaining their informed consent. In approaching the informants, the authors made use of their networks and mutual acquaintances to access the specific individuals who are in charge of or leading the digital and tech policy. For this exploratory study we focused on senior workers with organisations with a human rights background who are driving the agenda of digital rights and platform policy in the EU. The key informant interviews took place in March and April 2024, lasted between 60 and 90 minutes and were structured along three broad areas: (i) values, principles and organisational aspects; (ii) financial dynamics including funding and sustainability; and (iii) CSO stakeholder relations with platforms, policy makers, and other CSOs.

Adding auto-ethnographic insider perspectives

We added to these interviews auto-ethnographic insights where tensions and practical dilemmas are prevalent for Civil Society Organisations (CSOs) along with their associated pressure points. First, we delved into the values held by CSOs and individuals, seeking insights into the principles guiding interventions and policy contributions. Second, we explored funding, crucial for CSOs' independence and policy intervention capacity, examining direct or indirect involvement of platforms in funding, various forms of financial support available, and the impact of funding models on policy stances. We considered how policies and strategies are formulated and the measures in place to safeguard independence, delving into organizational structures and engagement policies with platforms and policymakers. Thirdly and finally, we examined other stakeholders in EU platform governance, including policymakers, notably the European Commission, platforms, and other CSOs, to understand the dynamics and relationships within the 'Brussels bubble' and the digital policy landscape. While each area was explored separately, discussions with informants often intertwined them, revealing connections between organisational values and processes, funding, and policy-making processes. Interviews were conversational, allowing informants to raise pertinent topics, enriching the insights gathered.

Public facing CSO document analysis

Our autoethnography and semistructured interviews were supported by an analysis of documents on our CSO informants organisational websites. This served to add depth and nuance, and to explain further the key issues. The document analysis presented the formal, public positions of CSOs, while we probed informants for the insider perspectives that do not necessarily become publicly articulated. Of the four areas we are interested in, values/principles and funding were most evident on the websites, while information on organisational policies and procedures was limited. We found no explicit references to relationships with other actors or parameters around stakeholder relations beyond proclamations of full independence from government. The document analysis allowed us to compare the public position of CSOs with the insider views of workers.

V. CSOs as platform policy actors: between revolution and reform

In this section we present the findings of this study structured along these three key areas: (i) values, principles and organisational aspects; (ii) financial dynamics including funding and sustainability; and (iii) CSO stakeholder relations with platforms, policy makers, and other CSOs. Although we posed questions on organisational factors and procedures, our findings suggest that these can be explained or understood through the prism of principles, political economic factors and relationships with other actors.

(i) Values, principles and organisational aspects

In approaching the values and principles and organisational aspects that underpin the work of the CSOs that participated in our study, we began with an analysis of their positions as stated on their websites. This analysis backed up previous findings, such as Celeste et al’s (2023) on the emergence of a convergence around certain key rights. Additionally, principles of independence, transparency and accountability were clearly articulated across all CSO websites we looked at. At a second level however, based on our autoethnographic understanding coupled with in depth interviews, a new ideological division emerged, which we can describe through the dilemma ‘reform or revolution’: while a ‘revolutionary’ position looks to dissolve and re-socialise platforms, a “reformist’ position retains platforms more or less as they are but introduces changes or improvements. We understand these positions as part of a continuum.

More specifically, our document analysis indicates that EU CSOs continue to speak broadly to human rights values in relation to the fundamental freedoms elaborated within treaties. CSOs articulate both specific and non-exhaustive rights, including privacy, protest, expression and assembly. Their public facing positions on rights like privacy remain influenced by conceptions of rights which predate the emergence of platforms (Zuboff, 2019). Public facing documents are consistent with Celeste’s (2019) argument outlining CSOs normative role as one of providing the basis for the creation of an emerging digital constitution conferring fundamental digital rights - in reference to long standing rights that predate digital concerns.

This focus on rights that predate platforms is neither unexpected nor controversial, showing a strong consensus among CSOs in terms of core values. Positions diverged however when it came to how these core values could be defended. It is here that the continuum of ‘reform versus revolution’ emerged. This is how our informants put it:

“I wouldn't waste any time on content moderation while we still have a massive [content] amplification problem. The companies are certainly not going to switch off the amplification systems until they're forced to. [...] The DSA inside the European Commission, they've set up a new priesthood with its own dogma and robes that everyone's going to wear, and they're spraying incense around the place. And it all looks very worthy. And the NGOs are writing their illuminated manuscripts, and it's going great right. Everyone has a role. And it's probably a complete waste of time. Instead of saying: Okay, it's pretty clear [the platforms] are just big utilities. Let's just take them over, or let's break them up, you know. Let's go big.” [Informant 1]

“My team is entirely focused on challenging the fundamentals of the business model of these platforms. That part of the conversation is by necessity quite adversarial and we certainly don't mince our words in saying that the entire business model conflicts with human rights, the right to privacy, and that we think it needs to be dismantled. The core of our work is a threat to their continued profitability and existence.” [Informant 4]

These informants take a ‘revolutionary’ position pushing for dismantling the current operational model of platforms. They reject both the policymaker efforts to regulate via the DSA and the platform self-regulatory efforts. Platforms are beyond reform and they are certainly not going to do anything out of their own initiative, while the European Commission lacks the will to properly regulate platforms as public utilities. CSO intervention should be oriented towards ambitious change and not lost in technicalities.

In contrast, reformist positions seek to identify positive CSO interventions to improve the way in which platforms operate. Reformist positions are also critical, but from a different perspective. For example, Informant 2 referred to their policy of asking for more transparent recommender systems, while Informant 3 raised the issue of harmed communities:

“We find it sometimes important to listen to [platforms] or know, not their strategies, but how they understand or what kind of problems they have when they argue that technically something is impossible to do. When we argue for transparent recommender systems, for example, and they argue that they cannot give transparency on recommender systems [...] [Informant 2]

“My strand of work was on sort of socialising the DSA. Figuring out what groups are going to be sort of ignored, depending on what policy? What policies are implemented. I'm very much interested in bringing questions of justice from an intersectionality framework into a framework like the digital services act. I would note: we're not talking about race, we're not talking about gender. We're not talking about options for redress around these issues like content, moderation, etc.. [State agents] were not interested, it was like, yes but it's more important that we have audits.” [Informant 3]

Reformist positions therefore assume a critical but pragmatic stance pushing for incremental changes that may improve platform governance, making it more inclusive, accountable and transparent. These positions shape and influence the approach and strategies these CSO workers pursue vis-à-vis platforms and policymakers. For example, CSOs with a reformist position engage with platforms directly:

“Our core mission is to advocate for the upholding of international human rights standards in tech policy, and because not all of that is appropriately done through strict regulation, we do direct advocacy to companies, so that they uphold their obligations to due diligence under the international human rights framework.” [Informant 5]

At the same time, one of our informants highlighted an important tension as, in the context of authoritarian countries, platforms can play an important role in supporting human rights defenders. While therefore the informant overall assumed a ‘revolutionary’ position they recognised that this involved a contradiction in practice because

“The monopoly power that these platforms hold is that there's so many human rights actors dependent on them basically to do their work, and we kind of sit in an interesting space where we do both the accountability work, a very strong critical accountability work. But also we also act as a kind of a conduit for those who need them.” [Informant 4]

This tension between the principled position of adversarial antagonism and the necessity of engaging with platforms emerged under a different guise across all areas we covered. We argue that this is the outcome of the neoliberal approach to regulation which prioritises self and co-regulation, while delegating important and necessary tasks to civil society without affording to it any powers to act or even in any way safeguarding its ability to act through appropriately resourcing it. This leads to the second area covered, the fraught issue of funding.

(ii) Financial dynamics including funding and sustainability

Financial independence and sustainability is a critical factor for civil society organisations. This is an ever delicate predicament for CSOs who have the dual purpose of advocating while managing funding flows. Our analysis indicates three categories of funding streams: membership, philanthropic foundation funding, and, to a lesser extent, project based funding from platforms. Regarding a membership funding model, where citizens supportive of the CSOs programmatic advocacy pay annual fees, the literature describes how this model confers a set of legitimacy regarding agenda setting and advocacy that the resource-bound donors lack (Neier, 2012). However, the nature of this fee structure can still be critiqued in relation to independence concerns given that the membership CSOs themselves are subject to critiques of hierarchical relations, while smaller CSOs do not have access to the same broad membership as larger CSOs. Tensions based on power imbalances and inequitable resource allocation are not resolved.

The second category of funding comes from large philanthropic foundation reliance. Large philanthropic donors have historically provided what is described as ‘core’ funding, enabling CSOs to deliver various programmes of work according to the CSOs’ mission, values and culture. This is a form of resource-bound donor funded reliance that, while maintaining independence from government, nonetheless creates a fragility for CSOs that may jeopardise their efforts to realise field-level impacts (Taylor, 2016). Indeed, there are examples in the literature of funding conflicts in this category: for instance, a large philanthropic organisation lessened funding flows when Human Rights Watch started a Women’s Rights Project (Neier, 2012). A further fragility attached to large scale philanthropy has been the shrinking of civil society through the gradual decline of core funding grants and the transfer to project based grants based on the specific missions and values of the funders themselves. Additionally, the large-scale philanthropic landscape has been identified as shrinking in Europe, with the funders pulling out, leaving EU CSOs scrambling to fill those deficits.

“The [funding] landscape in Europe is just shrinking, so we were extremely fortunate that we received a grant from a [large scale US based funder]. At the end of the year that funder is pulling out.” [Informant 5]

The final category of direct funding relationships, platforms, is more elusive. When asked whether they would be open to future funding relationships with platforms, CSO informants replied:

“No.” [Informant 1]

“Our funding policy prevents us from taking money from companies. But an issue is raised over and over again around whether we would be able to take money from big platforms for non rights related work, like rule of law or shrinking civic space. But the decision over and over again is that we don’t.” [Informant 2]

[There is] “an extremely strict policy position that we would never accept any corporate money from any corporation for profit corporation.”[Informant 3]

“That's a hard no.” [Informant 4]

“We haven't decided. Would we consider potentially taking, let's say, 10% of our funding from industry, as many other civil society organisations do [with the caveat that industry] do not get to dictate any of our policies, they don't have any influence on it.” [Informant 5]

All CSO informants were very clear that they did not receive direct platform funding. Document analysis shows how several of their CSO’s public facing documents stated explicitly that they don’t accept funding from industry with the others being silent on the matter. Our insider perspective reveals there is often a clear absence of written policy positions on the matter amongst EU CSOs. As some of our informants explained, the absence of platform funding, CSO resource-bound donor funded reliance and fragility may make them susceptible to various forms of financial influence. There were comments from our informants that other CSOs have this practice together with variations in positions towards future funding opportunities on a spectrum between a hard no to being open to strictly managed possibilities.

Another aspect of CSOs financial instability can be seen as reflected in their quest to carve a formal space for their interventions in favour of CSO programmatic longevity. Specifically, while there is a clear principled motivation to monitor the implementation of the DSA, economic self-preservation could be involved as well:

“But it became very clear that getting this section in the Digital Services Act, where there was transparency and auditing, was so clear because it gave civil society organisations a work remit. You could be like, well, we have access as a researcher, or we need access to data to do auditing your research. And that gives us a role to play [...] there is also self interest of civil society in part of the regulatory process [...] there's a push for transparency, for transparency’s sake …that I felt like was more about civil society trying to get a piece of the pie of the work rather than actually trying to push politicians or the EU or platforms to be better.” [Informant 4]

This precarity and quest for not only survival but also influence can be seen as underlying the relationships within the platform governance triangle which we discuss below.

(iii) CSO stakeholder relations with platforms, policy makers, and other CSOs

The key tensions observed here concerned platform power and power struggles for influence within the CSO sector. These tensions speak to the central problematic of neoliberal platform governance: as state/EU authorities prioritise a self and co-regulatory model through regulatory instruments like the DSA, platforms still have the upper hand, especially around problem definition and solutions; civil society organisations strive to act as watchdogs but without capacity, resources or definitional power, while also getting pulled towards different directions by the diverse priorities and power difference civil society organisations have.

Our informants offered considerable insights on the power platforms wield, highlighting the platforms’ efforts to set the agenda and drive issues in their own way. A crucial way of doing this is through exercising ‘soft power’ through stakeholder engagement events, leading CSOs to develop their own practices around these. Some CSOs will not engage at all, while others will try to set some rules for this engagement. These forms of indirect influence can be articulated from one of the authors’ own experience as a CSO worker, who received invitations to various meetings and events by platforms to discuss platform regulation advocacy efforts. CSO workers from major and EU influential CSOs were in the room. Platform representatives also included former CSO workers. Platforms provided research, communications, and advocacy information to aid mutually shared advocacy goals. However, attendance is not a consistent practice as some workers make decisions not to attend on a matter of blanket principle, see Informant 1:

“We don't trust them, and if we can avoid being in a room with them, we do”.

Or in an evaluative way on case by case basis, see Informant 3:

“We're constantly invited to round tables, policy discussions, consultations. And our take on that is, we will consider them on a case by case basis.”

Regardless, this norm of indirect capacity supports from platforms therefore seem to trickle down with direct and indirect impacts: (i) direct knowledge sharing could create indirect issue framing impacts; direct network development support lends to indirect alignment of a platform’s preferred allies; and direct communication support through CSO messaging amplification, comes with the indirect outcome of selective amplification of platforms’ preferred issues. These all speak to forms of platform capacity influence.

In parallel, platforms often co-opt civil society arguments to push their own agenda, especially around issues of freedom of expression:

“It was actually a push from Youtube […] that co-opted a civil society letter that was written …to almost make it seem like they aligned for civil society [...] and there had to be very difficult conversations to say, you cannot ever do that again, because that's not okay for you to do.” [Informant 5]

“And then we reached out to [large platform], asking them not to refer to us and claim that we have the same arguments because we didn't. So we try to explain to them where we are coming from, and even though the end result or the conclusion is the same - the liability of big platforms should be limited - we are coming from a different angle.” [Informant 2]

This apparent convergence of arguments is not entirely surprising since in liberalism both CSOs and platforms can be part of the private realm, standing against state efforts to limit freedom. Notwithstanding the fact that CSOs are oriented towards preserving the civil liberties of platform users and platforms are oriented towards profit, the end rhetorical result, as our informant put it, is the same.

Thirdly, platforms shift the discussion to technical matters which on the one hand challenge CSOs that have no in house technical expertise, and on the other make the discussion too complicated to be picked up by the media, interfering with CSO efforts to alert public opinion.

“There's so much to be decided on the implementation side. And it's no longer sexy for the media. So it they [platforms] have such an advantage because they're sitting waiting for every opportunity to leverage influence via lawyers, but also via their think tanks and their lobbyists, and it's harder and harder for civil society to kind of kick up a stink in opposition, because we're talking about such technical matters that don't make sense beyond the bubbles.” [Informant 3]

“Data analysis, I don't know a single civil society organisation that would understand transparency reporting of algorithms. [...] Knowledge, and especially technical knowledge, is missing from many, many NGOs.” [Informant 2]

More broadly, concerns expressed by our informants related to difficulties related to digital knowledge and technical understanding. They noted platform interventions into this space, with platforms indirectly recruiting CSOs focused on children, gendered and race rights that have vested rights interests, but lack parallel needed technical expertise, internal knowledge building capacity, or access to circles that are connected to and immersed in digital policy conversations. This results in divides between rights groups, some of whom don’t consider some policy ramifications of their interventions. Examples here include proposals for breaking end to end encryption to protect women: .

“[...] It's easy for a very large online platform to say, Hey, we're supporting civil society, because we've given X amount of money to this well known entity, but that entity may not have the necessary kind of nuanced technical expertise that they need at a policy level, not just from a tech policy level to understand what are the ramifications of some of the things that they're calling for. Online gender based violence is such a good example because platforms provided funding to women's rights organisations who will then say, but we should definitely break end to end encryption, to stop the spread of revenge, porn or deep fakes. But what about those domestic violence shelters that depend on end to end encrypted services, to speak directly to victims and get them out. The situations they're in of the women human rights defenders that we have to help escape from authoritarian regimes rely on end to end encryption.” [Informant 5]

This points to tensions that are internal within the field of civil society organisations active in this space. While digital rights CSOs have taken the lead in this, CSOs representing women’s rights and anti-racist groups are also increasingly involved. These latter CSOs can be better connected to policymakers because they have a longer history of engagement with EU policy making as well as formal mechanisms, such as the Gender Equality Strategy and the EU Anti-Racism Action plan, that makes it more likely for them to influence policy in specific ways. But the problem then is that, in combination with the decrease in funding for digital rights CSOs, this space is ripe for industry to engage in co-optation:

“My concern is the co-option of long standing, very traditional civil society organisations that have been established for a very, very long time with funding in a shrinking civil society space where huge funders have pulled out, especially from Europe.” [Informant 5]

But this is not the only kind of internal CSO tension. Our informants alluded to a space that can be murky in terms of funding relationships but also differing priorities and abilities to shape the debate. Here we can see the ‘reform versus revolution’ positions re-emerge. The ‘revolutionary’ position looks for power dynamics and expresses intense scepticism:

“I can’t actually speak to very strong specifics. But I still think it raises concerns around what would happen if it did come down to certain crunch questions, and what it means just to be coordinated by a group is just less radical in its overall analysis of the problems. What kind of questions are asked [...] The coordinator has a lot of power in deciding these are the top things on our agenda to discuss. And maybe if it was us, we pick different matters on the agenda, and if we're not discussing them at all, maybe big tech is free to influence those matters. So there, I'm talking about prioritisation of business model related concerns versus, for example, content, curation or content moderation related concerns. And that's speaking hypothetically.” [Informer 3]

In contrast, the ‘reform’ position is looking to create strategic alliances with other CSOs and policy makers in order to pursue its goals of platform policy reform:

"Only public interest actors, or human, those of us who are informed by the International Human rights framework that meet in the group that we have. And then we have something called the [redacted]. This is an event that we've done biannually, so we've done it under every [government change] and the [government official] has always supported the events that we've done, and we never invite industry to the events that we do. So, it's only CSOs, academics, public interest technologists and institutions, and Member States.” [Informant 5]

Although the sample of CSOs we looked at was small, fairly consistent positions have emerged in terms of their approach to platform governance. While none of the CSOs in our sample could be entirely identified with one or another end of the continuum, CSOs 1 and 3 could be placed closer to the ‘revolution’ end of the continuum, CSO 5 was closer to the ‘reform’ end, and CSOs 2 and 3 could be placed somewhere in the middle. Observing the divergence of positions even among organisations that are broadly aligned in their vision of digital rights, raises questions regarding the possibility of consensus and effective civil society interventions in platform governance. Additionally, differences in priorities and goals can lead to fragmented or conflicting interventions, such as for example the focus on the business model as opposed to the focus on content moderation. Finally, it is likely that reformist positions are favoured by both platforms and regulators, as they may be seen as more feasible and less threatening; however, this may come at the cost of achieving long term systemic change.

Overall, these tensions between platforms and CSOs and within the civil society sector foreground a chief problem in neoliberal platform governmentality: while states have delegated important policy tasks to the private sector, which includes both platform firms and CSOs, these are unequally equipped to participate. CSOs’ lack of resources and (knowledge) capacity can make them susceptible to platform soft power and to self-serving actions; at the same time, the CSO sector is diverse in terms of its ideological positions on platforms and the priorities it pursues. In this sense, CSOs are caught up in circular thinking around platforms, although the radical position seems to articulate an overall refusal to accept the terms of the debate as set up by platforms and the European Commission. The final and concluding section returns to the article’s research question and theoretical framework.

VI. Conclusion

This article sought to explore the question of how CSOs contribute to platform governance through analysing a set of in depth interviews with five key informants from leading EU digital rightsCSOs, who participate directly in shaping the digital policy landscape. The interview analysis is supported by autoethnographic observations and a website document analysis. Theoretically, we contextualised the role of CSOs within neoliberal governance, in which the triangle of state-platforms-CSOs is seen as reflecting the neoliberal political rationality of liberty for markets with limited state power. The normative role of CSOs is therefore to push against both regulatory and platform encroachment and overreach. To what extent are they capable of fulfilling this role? Our findings suggest that notwithstanding an overall convergence on the importance of upholding human rights, as indicated also by Celeste et al. (2023), digital rights CSOs operate across a continuum which we termed ‘reform versus revolution’, reflecting drastically different positions to how platforms should be regulated. In parallel, financial insecurity makes CSOs susceptible to indirect platform influence and may re-orient their practices towards self-sustenance. Finally, CSOs face important pressures from platforms, which seek to co-opt not only lines of argument but also CSOs active in adjacent rights areas without in-depth knowledge of the digital world. Therefore, while our article provides relevant insights into how digital rights CSOs relate to one another and to platforms, there is scope to consider further research into their stance towards state actors as the third “angle” on the platform governance triangle. State actors are not a passive background but active agents in governance that CSOs can navigate strategically. Further interview materials and autobiographical insights into this third angle could form the basis for future study.

Using Foucault's analytics of neoliberal governance we argue that advocacy for state/public matters, such as digital rights and protections, is delegated to civil society. Civil society is caught up in the contradictory position of ensuring that these concerns are articulated so that policymakers and platforms act on them, but in doing so they uphold the main parameters of this model of governance that has delegated to them effectively the role of sustaining digital rights without any resources and without a mechanism for reaching consensus among their diverse positions and interests. Although civil society actors are engaged in counter-conduct, or the refusal of the terms of this model of governance, some are still compelled to act on the preservation of at least some of the fundamental issues created by platforms, especially around harm and surveillance. And in this context, the precarious existence of even the most established CSOs, and tensions within the milieu, can undermine their ability to keep on doing their basic job let alone engage in a radical redefinition of the terms under which (platform) governance takes place.

Further, while our article seeks to be thought provoking for a general public seeking to follow platform governance processes, we would like to target specific comments regarding digital rights to CSO actors themselves. Given CSO’s unintended role in normalising platform activities, as tangled up with the power - soft and overt - of platform stakeholders in influencing CSO advocacy positions in the platform governance triangle, it may not be sufficient to simply refuse funding support from platforms on specific matters as many of our participants have described (and we recognise via our insider position just how challenging funding refusal alone can be in a time of significant platform economic influence and CSO resourcing precarity). Rather, stricter parameters might be drawn against soft influence around multilateral stakeholder methods on platform regulation matters. Platforms have the capacity to engage the state and CSOs sphere through lobbying, educational and indirect financial support through workshops and outreach. We have seen them host events, provide preliminary research on digital rights issues, and lead with narrative and framing examples. They fund educational programmes and infrastructure. They hire individuals who have experience and impact in public governance. In the face of this significant soft influence, digital rights CSOs could set terms of engagement around shared CSO values, principles and organisational policy and then adhere to them - loudly - as a principled example for our fellow CSO colleagues to follow.

However, and finally, CSOs must also be mindful of their own role in power hierarchies and their involvement in legitimation of practices that may end up hurting communities and exacerbating inequalities. We would therefore suggest that in addition to short term redrafting of their terms of engagement with platforms and regulators, they become involved in other parallel efforts to address deeper structural issues and inequalities.