Is Telegram a “harbinger of freedom”? The performance, practices, and perception of platforms as political actors in authoritarian states

ABSTRACT

This paper examines the practices, performance, and perceptions of the messaging platform Telegram as an actor in the 2020 Belarus protests, using publicly available data from Telegram’s public statements, protest-related Telegram groups, and media coverage. Developing a novel conceptualization of platform actorness, we critically assess Telegram’s role in the protests and examine whether Telegram is seen as playing an active role in Belarusian contentious politics. We find that Telegram’s performance and practices drive citizens to form affective connections to the platform and to perceive Telegram as an ally in their struggle against repressions and digital censorship. Meanwhile, the Belarusian state uses Telegram’s aversion to censorship and content moderation to intervene in contentious politics by co-opting grassroots approaches and mimicking manipulative efforts of other authoritarian regimes. Our conceptual framework is applicable to post-Soviet authoritarian contexts, but can also serve as a useful heuristic for analyzing platform actorness in other regime types.

KEYWORDS:

• Telegram
• platform actorness
• platform governance
• Belarus
• protests
• authoritarianism

Introduction

The idea that information and communication technologies (ICTs), and social media in particular, play an important role in contentious politics is firmly established. Yet, scholarship on contentious politics, especially regarding authoritarian states, insufficiently acknowledges and analyzes the agency of such digital platforms. Rather, platforms have been seen as primarily instrumental: operationalized as either the space where discontent and political participation happen (van Dijck and Poell 2015), or as the tools that enable citizens to mobilize against authoritarian rule (Valenzuela 2013; Skoric et al. 2016), while also providing regime actors with opportunities to suppress public discontent (Gunitsky 2015). These arguments also presuppose that citizens and/or regimes perceive and use digital technologies in an entirely rational and objective manner (e.g. based on their features). In reality, the reputation or public framing of a platform or technology can also shape how it is perceived and used.

In line with recent scholarship on platform governance (Gorwa 2019; Haggart 2020; Haggart and Keller 2021) and platform power (Helberger 2020; van Dijck, Nieborg, and Poell 2019), we argue that digital platforms themselves should be understood as actors with agency who participate in contentious politics and can shape the outcomes of political transformations. It is thus crucial to understand how a platform’s reputation is created, performed, and perceived by the various actors involved in contentious politics, and what the consequences might be if these perceptions are distorted or misrepresented. The latter dimension, in particular, has thus far received little attention from platform governance scholars.

To gain a better understanding of how and why certain tech companies come to be seen as “harbingers of freedom” and protectors of digital rights and how this impacts their role as actors in contentious politics, we examine messaging app Telegram’s performance and practices in the context of the 2020 Belarus protests against presidential election fraud, during which the Belarusian state sought to restrict protesters’ use of digital technologies, as well as perceptions of these actions by citizens and the state.

Telegram offers end-to-end encrypted messaging options, which is one reason it is popular among journalists, activists, and ordinary citizens seeking to shield themselves from governmental surveillance, in addition to public “channels” and groups. The company takes pride in resisting collaboration with state authorities, for which it has been banned in, for instance, Iran and (between 2018 and 2020) Russia (Akbari and Gabdulhakov 2019; Wijermars 2021). The combination of the platform’s affordances and comparatively lenient content moderation policies¹ have also made Telegram an attractive destination for extremist groups (Shehabat, Mitew, and Alzoubi 2017) and internet celebrities “deplatformed” by other services for violating platform policies (Rogers 2020). However, with regard to its use by those resisting authoritarian repressions (e.g. during the Belarus protests in 2020), the company’s position as an advocate of internet freedom tends to be uncritically accepted, especially within broader public debates (Herasimenka et al. 2020).

Large technological corporations often promote the idea that their activities serve to protect and promote internet freedom, political emancipation, and democratization around the globe (Jørgensen 2017). Yet, many of their corporate practices regarding content moderation actually harm societal interests and the interests of individual users (Jørgensen 2019; Nothias 2020; Viera Magalhães and Couldry 2020). Moreover, there is increasing evidence that platforms’ adherence to corporate responsibility principles is highly uneven (countries and regions are considered to be of high or low priority according to corporate interests) and PR-driven (Whitten-Woodring et al. 2020; Wong 2021). This is particularly concerning in authoritarian contexts where users are even more dependent on platforms and apps claiming to be “safe” from state interference and are vulnerable to significant real-life consequences should their trust in these companies be undermined.

By developing and applying a novel conceptualization of platform actorness, our paper aims to both critically assess Telegram’s role in the protests and to understand to what extent Telegram is seen as playing an active role in Belarusian contentious politics. This then allows us to assert to what extent its performance (i.e. how it frames its position and decisions) aligns with (the possible political impact of) its practices and with citizens’ and the state’s perceptions of its role – and its allegiances. We find that Telegram’s performance and practices drive citizens to form affective connections to the platform and to perceive it as an ally in their struggle against repressions and digital censorship. Meanwhile, the Belarusian state uses Telegram’s aversion to censorship and preference for unfettered speech to intervene in contentious politics by co-opting grassroots approaches and mimicking manipulative efforts of other authoritarian regimes. Users’ decisions thus rest on a balance of pragmatism and perceptions of risk, and Telegram’s reputation plays a clear role that requires further examination.

We first provide an overview of existing scholarship on the role of ICTs in global and post-Soviet contentious politics and offer a brief history of Telegram’s origins, its advocacy against state surveillance, and its activities supporting the circumvention of state-imposed limitations on internet freedom. Subsequently, we introduce our conceptual framework for operationalizing platform actorness and outline the data used for this study, before conducting an in-depth analysis of Telegram’s practices, performance, and perceptions in the Belarus protests. In our conclusion, we reflect on the applicability of our conceptual framework in other contexts, and outline directions for future research.

The role of ICTs in post-Soviet contentious politics

The democratizing potential of ICTs remains a highly debated topic in communication studies and political science. Special scrutiny is given to the potential of ICTs to enhance citizens’ capability for dissent and protest organization. In many post-Soviet countries, but also MENA states, Turkey, or Hong Kong, the internet and digital platforms themselves have become highly contested political spaces in which the state and civil society are battling for control (MacKinnon 2012; Rane and Salem 2012; Lee, So, and Leung 2015; Tufekçi 2017). Scholars find that in many post-Soviet countries the internet is perceived as a “liberator” and as an alternative to traditional media co-opted by the elites, although this perception is weighed against the repressive capabilities of many post-Soviet regimes (Fossato, Lloyd, and Verkhovsky 2008; Oates 2013; Gunitsky 2015; Pearce, Vitak, and Barta 2018; Wijermars and Lehtisaari 2020a).

The liberation potential of online platforms in the post-Soviet sphere was long predicated on the relative deregulation of the internet compared to traditional media. The last decade has seen a significant growth of more formal internet controls across former Soviet states, including a wave of repressive internet regulations specifically aimed at citizen expression and activity online (Pearce and Kendzior 2012; Anceschi 2015; Maréchal 2017). In Russia, even before state control of the online space was tightened, online political discourses “were actively manipulated by authorities, a strategy that has since been exported” to and emulated by other post-Soviet states (Wijermars 2021). Analyzing the evolving “illiberal” practices of internet control across the post-Soviet region, Kerr finds “significant areas of similarity, with evidence both of uncoordinated diffusion, learning, and emulation, as well as direct collaboration or coercion mechanisms driving this convergence” (Kerr 2018, 3814). For example, SORM, Russia’s technical surveillance tools, have at various times been implemented by Ukraine, Belarus, Kazakhstan, and Uzbekistan, while Kyrgyzstan, Tajikistan, Turkmenistan, and Azerbaijan have made strides toward developing or acquiring similar systems for lawful interception of telephone and online communications (Kerr 2018, 3822).

Across the region, ICTs have played an important role in activist movements (Beisembayeva, Papoutsaki, and Kolesova 2013; Gobert 2013; Oates 2013; Pearce, Vitak, and Barta 2018), as well as more prominent large-scale protests (Goldstein 2007; Enikolopov, Makarin, and Petrova 2020; Lonkila, Shpakovskaya, and Torchinsky 2020; Lokot 2021). However, as Oates (2013) notes, while ICTs may offer the same opportunities when regarded in isolation, national media and political systems play a key role in shaping and constraining protesters’ perceptions and affordances of digital platforms.

The tendency to view internet platforms as alternatives to mainstream, state-controlled media is bolstered in many post-Soviet states by the fact that most of the top social media networks are run by foreign corporations. With the notable exception of VKontakte (owned by Russian MailRu group and popular among Russian-speaking users), the dominant majority of digital platforms are owned by Google, Facebook, or other Western companies. For post-Soviet internet users, this Western orientation often signifies that these platforms may be more concerned with free expression and user rights and have more developed privacy settings (Vasilyeva 2012).

Perceptions of platform credibility and responsibility are also shaped by public awareness of formal and informal platform governance (Gorwa 2019; Haggart 2020), public debates about the platforms’ responsibility to uphold international human rights principles (Jørgensen 2017, 2019), and visible evidence of platform support for internet freedom and digital rights (Helberger 2020). In recent years, the role of (global) platforms has been increasingly questioned in both democratic and non-democratic states amid concerns about user privacy and the spread of disinformation. However, platforms’ reputation as harbingers of freedom persists in the coverage of their role in protest events. With regard to Belarus and, earlier, Russia and Hong Kong, mainstream media coverage has been dominated by reports extolling the virtues of Telegram as the “secure app” and dubbing the protest events “Telegram revolutions” (Baunov 2020; Williams 2020; Lynch 2021), resonating with the by now discredited claims about Twitter and Facebook revolutions of the 2000s and 2010s.

Media framing also plays a key role in shaping public perceptions of platform credibility. Survey research from Russia, for instance, indicates that citizens who rely on state-controlled federal TV news perceive the internet as a threat more often and indicate higher support for political censorship online (Nisbet, Kamenchuk, and Dal 2017). On the other hand, there exists significant criticism of activities of (Western) technology companies in the so-called Global South (e.g. Biddle, Finch, and West 2017; Nothias 2020), including many authoritarian regimes, which juxtaposes digital platforms’ performative corporate responsibility rhetoric with their relative lack of action in countries that are less prominent on the global stage.

This study contributes to scholarship on ICTs and contentious politics in post-Soviet states by proposing that digital platforms be examined not only as tools of political participation or mobilization, but also as actors in their own right. We suggest that their claims, performances and practices in the context of civic and political activity should be closely studied, along with how their “actorness” is co-constructed by state actors, media, and the public. Doing so allows us to gain a fuller picture of the dynamic unfolding of protest events and helps to better explain user motivations and patterns of platform use. In this regard, Telegram’s role in the Belarusian protests presents an interesting case for analyzing platform actorness as it is an arguably independent actor external to Belarus that became a central part of the public discourse of the protest and has heretofore been outspoken about digital rights and user privacy. Despite a history of state repressions against civil society and independent media, Belarus has received significantly less attention in scholarship on ICTs and contentious politics, and has been overlooked as a site of potential accountability for digital platforms operating in authoritarian regimes.

We therefore begin by considering the evolution of Telegram in the context of political histories of platforms and how their ideology and politics have been shaped by state approaches to governing platforms and policing networked spaces. An in-depth analysis of Telegram in Belarus, in particular, then is presented in a following section (Telegram as an actor in the 2020 Belarus protests).

Telegram: a brief history

Telegram was founded by Russian tech entrepreneur Pavel Durov and his brother Nikolai in 2013 and is currently one of the most popular messaging services in the former Soviet space, with over 500 million active users in January 2021 (Schroeder 2021). Telegram’s founders had previously founded Russia’s most popular social networking site VK (VKontakte), with Pavel Durov widely known as the public face of VK. In 2014, Durov “was forced to sell his shares and resigned as the company’s CEO following his continued refusal to cooperate with the Russian authorities by sharing user information” (Wijermars and Lehtisaari 2020b, 6). Durov left Russia in 2014, and announced Telegram’s launch soon thereafter.

The messaging app emerged in the context of the Kremlin’s growing crackdown on civil society and political expression in Russia, and is widely believed to have been “created as a tool to protect their communication in the context of political persecutions” (Ermoshina and Musiani 2021). In its public-facing messaging, the platform highlights support for internet privacy as one of the key tenets of its mission and states that it seeks to protect “your private conversations from snooping third parties, such as officials, employers, etc.” and “your personal data from third parties, such as marketers, advertisers” (Telegram n.d.). At the same time, the platform adopts a unique approach to encryption: while messages in its Secret Chats use end-to-end or client-client encryption,² regular Cloud Chats use client-server/server-client encryption and are stored encrypted on Telegram’s own cloud servers. Telegram argues that this “allows Telegram to be widely adopted in broad circles, not just by activists and dissidents, so that the simple fact of using Telegram does not mark users as targets for heightened surveillance in certain countries” (Telegram n.d.). This approach, where end-to-end encryption is not enabled by default (and not at all available for group chats), and the use of its own proprietary encryption protocol have made Telegram the target of intense criticism from tech experts and digital rights activists, who argue that the platform misleads users with overblown claims about its security (Greenberg 2021).

However, as Maréchal (2018) notes, the platform’s politics have also been shaped by Durov’s own cyber-libertarian views and by Telegram’s peculiar choice of business model, which until recently has relied predominantly on venture funding (Durov announced in December 2020 that Telegram would begin to monetize its platform by offering advertising in public channels and premium features, though personal conversations would remain ad-free [Durov 2020b]). As Maréchal’s analysis shows, Telegram’s public rhetoric “emphasizes user security, privacy, and freedom of expression,” but the company has so far failed to demonstrate in any tangible or verifiable ways that it lives up to these commitments. Unlike some of its competitors, the messaging platform has not resorted to concrete transparency and accountability measures, and seeks to base its credibility on its users’ blind trust in its good intentions and the platform team’s stated credentials.

Telegram’s functionality and stated values have proven to be especially popular with activists and dissidents, from Iran and Hong Kong to Russia and Belarus, who appreciate its relative anonymity and security, its efforts to remain accessible in the face of state censorship, and its lax approach to content moderation (Scollon 2021). But the same reasons have also earned it a loyal following among extremist and terrorist groups, from ISIS to the far right, and other shadowy groups seeking to coordinate their actions and share information out of the limelight (Meineck 2020). This diversity of Telegram publics has meant that in Russia, for instance, Telegram now occupies an important space in the public sphere, hosting various channels sharing political commentary, leaks and digital rights advice, and is equally popular with opposition activists, journalists, and Kremlin officials (Lokot 2018b). But the presence of less savory groups on the platform has also given the Russian state a convenient reason to outlaw the app, which it attempted to do in 2018 by blocking Telegram, albeit unsuccessfully (Wijermars 2021).

The recent moments of mass mobilization and digital repression in Belarus offer a useful opportunity to examine how Telegram and other actors co-construct the platform’s actorness and credibility and how they ascribe political agency to Telegram in the context of Belarusian contentious politics. Analyzing these overlapping and competing representations, we also seek to critically examine whether the platform’s practices align with its performative digital rights advocacy, and the implications that possible gaps between practices and performance might have for Belarusian citizens who rely on the platform in moments of mass mobilization and for the Belarusian state which, in turn, seeks to co-opt or subvert such mobilization potential.

Approach

Conceptualizing platform actorness in contentious politics

We depart from the assumption that how users engage with digital technologies during times of political contention derives from a weighing of functionality (affordances; audience) and assumed risks. The reputation of and public discourses surrounding particular services, in turn, may (sometimes incorrectly) inform the assessment of the latter. Here, we build upon scholarship on how activists in post-Soviet states negotiate visibility and security in their online resistance practices (Lokot 2018a; Pearce, Vitak, and Barta 2018).

To analyze platforms as actors in contentious politics, we propose a conceptual framework of platform actorness comprising three dimensions: practices (enacted actorness), performance (framed actorness), and perception (perceived actorness). Below, we unpack these dimensions and indicate how they may (help) shape unfolding protest events.

Platform practices

The term “platform practices” refers to the actions and decisions taken by a platform that circumscribe whether and how users engage with the platform and its features, and can potentially impact on contentious political events. We differentiate between four categories of practices.

Technical facilitation of use: Actions by the platform to facilitate access for users or to enable use in circumstances where such access or use might be restricted. Such measures can involve creating or promoting censorship circumvention tools (such as virtual private networks [VPNs] or proxies) or alternative means of access, such as website mirrors. Such actions may affect access to and sharing of information, including for mobilization and coordination purposes, and are of particular importance in case of government blocking of websites or internet shutdowns.

Content moderation: Decisions a platform makes with regard to removing or blocking user accounts or content, often at the request of (repressive states, and whether the platform’s content moderation policies are designed to prevent/enable content takedowns or account removals. Since platforms have the capacity to act as an arbiter on the accessibility of particular content, accounts, or groups and, thereby, shape online expression, their content moderation decisions have political salience. At times of political contention, they are also often interpreted as being political (President Trump’s Twitter ban being a prime example). While (over)moderation receives the most attention, we argue that refraining from moderating (undermoderation) should equally be seen as a political act.³

Data sharing: How a platform stores and protects users’ data and how it makes decisions about sharing that data with third parties, including states and law enforcement. Data sharing by platforms poses a risk to individual users, who may be identified and targeted/prosecuted for their (online) statements or activities, as well as to groups and movements organizing against repressive regimes.

Platform design and affordances: How the platform makes changes to its design, features, and affordances in the context of contentious politics. Such changes can be technical (e.g. facilitating anonymous or pseudonymous access), communicative (adding new features that afford new forms or levels of communication), or symbolic (non-technical changes to the platform that do not engender new features but provide largely symbolic value, e.g. changes to color schemes, new graphic elements). Actions falling within this category can affect communication dynamics and the forming and consolidation of (virtual) communities, and shape how users can express themselves on the platform.

Platform performance

The term “platform performance” refers to how platforms communicate their actions and values in the context of contentious politics. Through the framing of particular decisions or the rationales provided for releasing new features, platforms seek to shape public perception of these and maintain their corporate image. For research purposes, it is important to, first, be attentive to how corporate communication develops with a particular audience in mind and/or in response to public debate. Second, we should also aim to identify those actions and decisions that are not publicly commented on or explained by platforms. Studying how platforms “perform” their actorness is important as users have limited insight into how platforms work (e.g. regarding moderation practices). Publicly available statements concerning corporate values and key decisions can therefore be assumed to shape users’ understanding of the platform’s “operating logics” and suitability to the users’ aims and personal situation, along with other sources of information such as the experiences of peers. As platform reputation factors into users’ risk assessment, misleading or incomplete platform performance may obfuscate vulnerabilities that arise from using it.

Platform perception

The term “platform perception” refers to how a platform’s actions and performances are perceived by various actors and audiences. These perceptions are of importance to contentious politics in different ways. How states perceive platforms’ actorness informs whether they see the platform as a threat and what types of regulatory action they may take to influence or restrict its operations. It also informs assessment of whether and how a platform can be used to a regime’s advantage and the possibilities for interfering with, for example, coordination of protest events. How participants in contentious politics perceive a platform’s actorness informs their decisions on whether or not to use it, what they can (safely) say or share on the platform, as well as their assessment of the reliability of information shared there and the identity and intentions of other users (to what extent the platform shields against manipulation and surveillance). Finally, perceptions formulated in domestic⁴ media coverage can be seen as an important source for both state officials and citizens.

Methods and data

We collected publicly available data and materials (statements, posts, and reports/articles) relating to the Belarusian case generated by: Telegram and its employees/managers; the top national media covering the events in Belarus and Telegram’s role in them; and various stakeholders who interact with Telegram or frame its actions in their own statements (government officials, digital rights groups, and Belarusian citizens). Though not by design, most of the data collected is Russian-language content, with the exception of Telegram’s public posts and statements, which were available in both English and Russian. In our analysis, we combine document analysis with qualitative content analysis, tailoring our approach to each particular data source as is specified in the sections below.

Time bounding

We collected data for the period July–September 2020, which accounts for the earliest and most pronounced wave of protest mobilization in Belarus and the ensuing state crackdowns on civic and digital freedoms, including internet shutdowns and attempts to block or pressure social media platforms. The Belarusian data are placed in the context of earlier salient statements and events, where relevant.

Data sources

The data and content used in our study was drawn from the following sources:

Platform statements

We examined all publicly released posts or statements on Telegram’s official channels (blog, official Telegram channel, official Twitter account) and on founder Pavel Durov’s personal channels (Telegram channel; Twitter) for the period of 1 July–30 September 2020, reviewing them for any mention of Belarus, internet shutdowns, censorship, privacy, or anonymity.

Media coverage

We examined the online archives of four of the top five Belarusian media websites according to the “Media in Belarus 2020” report (Belarusian Association of Journalists 2020): NavinyBy (naviny.online, independent), Komsomolskaya Pravda Belarus (kp.by, Russia-owned), Belta (belta.by, state-run), and Belsat TV (belsat.eu, independent, based in Poland) for mentions of the company’s statements, actions, and interactions with Belarusian state authorities, as well as its use by protesters and the state, using keyword searches for “Telegram” (in Latin and Cyrillic script). Online news archives for the top news website in Belarus, tut.by, were not available as they were removed in June 2021 after the website was blocked by state authorities and many of their journalists arrested (Meduza 2021). Media coverage containing relevant keywords was collected for the period of 1 August–30 September 2020, when key internet shutdowns and platform censorship attempts occurred during the state’s crackdown on protests. The resulting dataset included 52 articles from kp.by; 386 articles by belta.by; 4 articles by naviny.online; and 83 articles by belsat.eu.

Statements by state actors

We noted any relevant statements framing Telegram’s actions in Belarus by Belarusian state actors (e.g. government officials or spokespeople) in the domestic media coverage for the same period.

To examine citizens’ perceptions of Telegram, we relied on a dataset⁵ from the largest and most active local Telegram group chat for Minsk, the Belarusian capital, affiliated with the Belarusian 97% movement that emerged in advance of the 2020 presidential elections (Herasimenka et al. 2020). The dataset comprises all conversations from this public group between 22 June and 23 August 2020. Non-textual elements, such as images, video, and stickers, were not captured. For close analysis, we focused on chat conversations on 9–10 August 2020 to capture the first wave of protests and ensuing repressions and internet shutdowns, when the number of new users in Telegram channels and the number of conversations grew considerably (Shelest 2020). The group chat conversations (N = 15,755) were manually assessed for mentions of Telegram and rationales for (not) using it or other perceptions of the messenger (regarding anonymity; risk of manipulation; surveillance; content moderation) presented by users. We also performed targeted searches on the full dataset of the group chat using relevant keywords (e.g. “Durov”, “phone number”).

Telegram as an actor in the 2020 Belarus protests

Background

Mass protests began in Belarus on 9 August 2020 when the long-standing president, Alyaksandr Lukashenka, claimed victory in the Presidential elections, with the Central Election Commission announcing he had won 80.1% of the vote (Central Election Commission 2020). This figure was challenged by opposition candidate Sviatlana Tsikhanouskaya, and the election results led to mass expressions of discontent by Belarusian citizens. Uniquely for Belarus, the protests were not centralized in the capital city, Minsk, but covered over 100 cities and towns by the end of the first week of discontent (Mateo 2022), especially following violent dispersal of protesters by state security forces, who made over 7,000 arrests in the first week (Human Rights Watch 2020).

Early analysis of the protests showed that, given Belarus’s heavily state-co-opted media landscape, protesters relied on independent (largely online) media (Greene 2022) and informal communication channels (Onuch 2020). Social media platforms were used widely for organizing, coordination, and information sharing by protesters (Shelest, Shutov and Nazarenko 2020; Gabowitsch 2021), though were less central to initial mobilization of cross-cleavage protest publics (Herasimenka et al. 2020; Douglas et al. 2021). Much of the popular attention in this context has focused on Telegram, though, as Gabowitsch (2021) notes, its ascent to prominence was unexpected, as in late 2019 it ranked as only the fourth most popular social media app in Belarus, a market that has been heretofore dominated by Viber, another messaging app developed by a Belarusian boasting a 70% market penetration in early 2020 (Gabowitsch 2021).

In attempting to explain Telegram’s sudden popularity, Gabowitsch (2021) suggests several reasons why Belarusian users flocked to the app, including their perception of Telegram as being more secure (compared to, e.g. Viber); Telegram’s broader use by protesters globally; the symbolic impact of Telegram’s founder Durov’s overtures towards Belarusian protesters; and, finally, the perception of Telegram as a platform specifically for protest activity vs. Viber’s predominant use for everyday conversations in Belarus. Building on this last explanation, Gabowitsch concludes that Telegram facilitated “a type of collective action dominated by the civic regime of engagement” (2021, 33). However, this conclusion limits Telegram’s actorness in the protests and presents an instrumentalized, non-agentic view of the platform in Belarusian contentious politics. In our analysis, we foreground Telegram as an actor and examine the key dimensions of its actorness, as enacted through its practices, framed through its performances, and as perceived by the other actors involved. The presentation of our findings follows the structure of the conceptual framework introduced above.

Practices: Telegram’s enacted actorness

Technical facilitation of use

While Telegram did not introduce new means of facilitating access for Belarusian users during the protests, there is evidence that it deployed the mechanisms created earlier to circumvent Telegram’s blocking in Russia in 2018. These included “relying on third party cloud services to remain partly available for our users [in Russia]” and “giving out bitcoin grants to individuals and companies who run socks5 proxies and VPN”; actions that Pavel Durov referred to as “Digital Resistance” (Durov 2018). On 22 June 2020, after the Telegram ban in Russia was lifted, Durov again referred to the technical supports on his official Telegram channel, writing:

In April 2018, Russia’s telecom regulator Roskomnadzor blocked Telegram on the country’s territory. We knew it was coming, so by the time the block went live, we had already upgraded the Telegram apps with support for rotating proxy servers, ways to hide traffic and other anti-censorship tools. We were joined by thousands of Russian engineers that set up their own proxies for Telegram users, forming a decentralised movement called Digital Resistance. (Durov 2020a)

Despite the Russian standoff ending with “a ceasefire,” Durov said the Digital Resistance “unblocking technology” was being regularly updated and would be used to facilitate access to the messenger in other countries, such as Iran and China (Durov 2020a).

In a Twitter post on 10 August 2020, as the Belarusian protests gathered steam amid ongoing internet shutdowns (Netblocks 2020), Durov revealed that Telegram had deployed its technology to facilitate access for Belarusian citizens:

We enabled our anti-censorship tools in Belarus so that Telegram remained available for most users there. However, the connection is still very unstable as Internet is at times shut off completely in the country. (Durov 2020c)

Despite unstable connections, Telegram remained largely available to users in Belarus throughout August and September 2020, and analysis of Telegram protest groups at the time found that both membership and message numbers increased significantly (Shelest 2020). At the same time, Telegram’s intervention also enabled continued access to the chats for Belarusian law enforcement, allowing them to spy on protesters. It also facilitated the creation of state-run Telegram channels disseminating counter-protest messaging. Incumbent president Lukashenka’s press service created its own channel, Pul Pervogo (“The First One’s Pool”), a carbon copy of challenger Sviatlana Tsikhanouskaya’s channel Pul Pervoy, to post regime-friendly videos and updates (Scollon 2021).

Content moderation

Telegram favors a hands-off approach to content moderation, and states in its FAQ that it processes only “legitimate requests to take down illegal public content (e.g. sticker sets, bots, and channels)” (Telegram n.d.), but notes this does not apply to private chats or “local restrictions on freedom of speech”:

For example, if criticizing the government is illegal in some country, Telegram won’t be a part of such politically motivated censorship. This goes against our founders’ principles. While we do block terrorist (e.g. ISIS-related) bots and channels, we will not block anybody who peacefully expresses alternative opinions. (Telegram n.d.)

Durov himself has argued that removal of terrorist content by Telegram “is consistent with our values” (Durov 2019a) and suggested that while Telegram supported “free speech and peaceful protest,” “terrorist propaganda has no place on our platform,” noting that users “don’t have to sacrifice privacy for security” and can “enjoy both” (Durov 2019b).

In the Belarus case, we have not been able to identify instances of moderation decisions that were challenged by protesters. A key content moderation dilemma emerged in late September–early October 2020, when Telegram revealed that Apple (acting on a complaint from the Belarusian authorities) had asked them to take down three Telegram channels engaging in doxxing (i.e. exposing personal information) of Belarusian law enforcement officers. On 8 October 2020 both Telegram’s Belarus-specific channel (created in mid-August 2020) and Durov’s own channel announced Apple’s request to shut down the three channels in question (Telegram Belarus 2020a; Durov 2020d). Durov stated that he was in favour of keeping the channels open, but noted that Apple would likely block these channels for iOS users even if Telegram refused to block them (Durov 2020d). We will return to how Telegram framed this decision and sought to engage Belarusian citizens on this question below.

Our content analysis of the “97% Minsk” chat group revealed another instance where Telegram did not moderate content, while other platforms may have chosen to do so. On both 9 and 10 August, we identified repeated discussions about Molotov cocktails, including the sharing of instructions for making them. Such information typically falls within the scope of extremism and/or incitement to violence, and thereby can be grounds for removing the post or blocking an account, group, or channel for most mainstream platforms. While it is beyond the scope of this study to determine whether the suggestions to use Molotov cocktails and instructions for making them may indeed have originated from “provocateurs” (as participants in the chat claim), this should then be seen as a double provocation: inciting real-life escalation of events and acts of violence (in turn, justifying increased repression and undermining peaceful protest efforts) and provoking the disruption of information-sharing by seeking to set in motion moderation mechanisms.

Data sharing

Telegram’s overall approach to guarding its users’ data is demonstrated in its approach to encryption, as well as its policy on requests for user data. As stated earlier, Telegram only enables end-to-end or client-client encryption for its Secret Chats, which “leave no trace on our servers, support self-destructing messages and don’t allow forwarding” (Telegram n.d.), but are not turned on by default. Regular chats, group chats, and channels use client-server/server-client encryption and messages are stored encrypted on Telegram’s own cloud servers, although public groups and public channels “will be accessible to everyone” (Telegram 2021). According to Telegram:

This means that all data (including media and files) that you send and receive via Telegram cannot be deciphered when intercepted by your internet service provider, owners of Wi-Fi routers you connect to, or other third parties. (Telegram n.d.)

Telegram claims it does not disclose the content of private chats or group chats to third parties and does not process “any requests related to them” (Telegram n.d.). However, Telegram concedes that user data and account information are still vulnerable to discovery through “physical or root access to your phones or computers” (Telegram, n.d). Critics also posit that Telegram is less forthcoming about the fact that it collects and stores user metadata (IP address, devices used, history of username changes) and a list of users’ contacts, if the user chooses to share them; there are examples of successful state attempts to extract such data, as well as the content of channels and group chats (Meineck 2020). While Telegram mentions in its privacy policy that it collects this data, it does not specify if it is encrypted when stored on its servers.

There was no direct evidence of Telegram receiving any requests to share Belarusian users’ data with the state or law enforcement during the protests. However, several public Telegram channel admins were arrested in June 2020, and many others were identified, apprehended, or persecuted later on charges of instigating “mass riots” (Scollon 2021). There is also extensive evidence of police confiscating protesters’ phones and attempting to discover whether Telegram was installed and to gain access to their subscriptions and contact lists (Kurochkina 2020).

Platform design and affordances

Within our chosen timeframe, Telegram introduced several features and updates that are relevant to how Belarusians were able to express themselves on the platform, but also some that reveal a political alignment with the protesters. On 15 August, the platform introduced a dedicated Belarus channel (https://t.me/s/telegrambelarus/). While used infrequently, it signaled Telegram’s prioritization regarding the events in Belarus. On the same date, the Belarusian language was made available to all users, along with a collection of themed stickers with the white-red-white Belarusian flag. While the former can be categorized as a communicative change, the introduction of emojis to express the flag is symbolic as its colors have come to represent the protests and any use of white and red is seen as subversive by the regime. The release is therefore indicative of Telegram taking a particular stance with regard to the protests and framing their other actions in this light.

In a 30 September 2020 update, Telegram introduced two new features. One new anonymity-focused setting called “Anonymous Admins” allowed group administrators to remove their names/pseudonyms and to become invisible in the list of group members. Telegram noted that messages from newly anonymous admins would “show no trace of their personal account and instead are signed with the name of the group” (Telegram 2020). Another new feature enabled users to add comments to posts in channels by introducing a comment button under each post (Telegram 2020). While the admin anonymity was a technical change, the chat comments also added new communicative affordances, allowing easier discussion in group chats.

Another Telegram feature popular among protest participants and activists is the ability to obfuscate users’ phone numbers that were previously visible as part of a user’s account profile to people in their address book (Telegram n.d.). Telegram added an option to hide phone number information during the Hong Kong protests, but the feature was also extensively referenced by Belarusian Telegram chat participants in the conversations in our sample.

Performance: Telegram’s framed actorness

Both Telegram’s official platform channels and Durov’s personal channels were used for public framing of the messenger’s actions during the Belarus protests. While performance regarding the practical aspects of enacted actorness, such as facilitating access to the app or decisions about data sharing, was limited, there were several instances of performative framing of other types of activity relating to content moderation decisions and changes to platform design and affordances, especially with regard to security and privacy. Both Durov and Telegram made several specific references to the situation in Belarus when framing the platform’s actions to enable access and fight censorship, while in other cases their statements alluded to it only indirectly.

Though there was limited performative activity on Telegram’s part during the early days of the Belarus protests (beyond announcing its circumvention tools had been enabled), in a 22 June 2020 post published in the wake of Russia lifting the Telegram ban, Durov referred to the Digital Resistance tools as “unique among social media apps” (Durov 2020a). Durov explicitly framed Telegram as a “privacy-focused” app and stated that the technology facilitating access to the app was being continually updated for use in other repressive states:

We don’t want this technology to get rusty and obsolete. That is why we have decided to direct our anti-censorship resources into other places where Telegram is still banned by governments—places like Iran and China. We ask the admins of the former proxy servers for Russian users to focus their efforts on these countries. They should also stand ready for new challenges: as the political situation in the world becomes more unpredictable, more governments may try to block privacy-focused apps like Telegram. (Durov 2020a)

With regard to the content moderation decision stemming from Apple’s demand to shut down three Telegram channels exposing the personal data of Belarusian law enforcement, Telegram and Durov framed the channels in question as “used by the people of Belarus to expose the identities of their oppressors,” “punishers and propagandists” (Durov 2020d; Telegram Belarus 2020a). Durov further framed the situation as ambiguous, saying that in his view “this situation is not black and white” and that he “would rather leave the channels be” (Durov 2020d). In his follow-up post a day later, Durov framed Telegram’s decision not to shut down the channels as stemming from popular opinion, reporting that according to a Telegram Belarus poll (Telegram Belarus, 2020b), “over 94% of Belarusian users think the channels that made Apple worry should be left alone” (Durov 2020e). The decision to frame individuals whose data were compromised as “oppressors” and the pretense that Telegram’s views aligned with the will of the Belarusian people indicate that Telegram had clearly chosen to side with the protesters in this ethical debate. Its framing shows that its “privacy-focused” approach is as selective as its undermoderation, while allowing the platform to retain its stance as a proponent of free speech.

Telegram’s performance of its actorness was most obvious in how the platform and its founder framed new features and design changes. In its 30 September blog post announcing the features enabling administrator anonymity and channel comments, Telegram explicitly drew a direct connection with the events in Belarus, linking to a Radio Free Europe/Radio Liberty report about the use of Telegram in the protests. The platform framed its design changes in terms of its core stated value of free speech:

Today’s update also features Anonymous Admins to make your battle for freedom safer and Channel Comments to make sure everyone’s voice is heard.

Telegram is increasingly used to organize protests for democracy and freedom. Today we’re introducing another tool for safer protests. (Telegram 2020)

Commenting on the Anonymous Admins feature, Durov obliquely said on his own Telegram channel that it was “a handy feature for celebrities and generally for anyone who wants to run a large group chat without publicizing their personal account” (Durov 2020f).

The creation of the Belarus-specific channel and other more symbolic gestures such as the creation of a new sticker pack in support of the Belarusian protesters were not referenced by either the Telegram channel or Durov’s channels. However, their creation is in itself a performative act that contributes to Telegram’s framing of its stance and alliances with regard to the protest.

It is evident that in performing its actorness, Telegram frequently references its mission as a secure app concerned about user privacy and frames its features and tools as enabling free expression and resistance. It deliberately presents itself not as a neutral technology, but as one with a specific positionality and repeatedly alludes to the values invoked in its policies and public statements.

Perception: Telegram’s perceived actorness

Media coverage and perception

Overall, Belarusian media coverage did not reveal many direct references to Telegram as an actor actively engaged in the Belarus protests. However, we observed some indirect framing trends and notable differences in how independent media and state-run or state-friendly media framed Telegram’s role.

Independent media outlets NavinyBy and Belsat TV predominantly framed Telegram’s role in the protests in instrumental terms, discussing the role of specific Telegram channels and how they were used by protesters to organize, share information, and coordinate protest activities. Both media outlets also featured episodic coverage of repressions against protesters, which included instances of police confiscating their phones to access their Telegram profiles and subscriptions to reveal channel administrators. Importantly, both NavinyBy and Belsat TV reported on internet shutdowns in August and September 2020 and Telegram’s halting, but continued accessibility throughout the period.

On 29 August 2020, NavinyBy featured an interview with Stsyapan Putsila, editor of NEXTA_Live telegram channel, a key source of protest-related information that also played a coordinating role. Putsila was quoted as saying that ICTs were now “not the fourth, but the third or second estate” (NavinyBy 2020a), though he did not directly assess Telegram’s role in the protests as a platform.

NavinyBy and Belsat TV both reported on Telegram introducing channel comments and on Durov’s remarks regarding Apple’s request to take down channels for doxxing law enforcement officers, but did not frame them in any particular way.

Belsat TV predominantly published protest-related news stories mentioning various Telegram channels as sources, including protest-related ones and state-run ones. Its coverage mentioned protest channels calling for protest rallies in specific locations; Telegram users joining the protest after calls in Telegram channels (with the police claiming they were paid to do so); or users arrested for making Molotov cocktails after allegedly finding instructions in Telegram channels. A 21 August news report about a young protester detained and beaten by police in the town of Mahilyou cited the victim as saying he had met up to protest with other people from a “Telegram chat” and mentioned that among them were police officers who knew about the rally having infiltrated the chat (Belsat 2020a). In another piece from 7 September, a student protester relayed that police used the presence of Telegram on her phone and her subscription to several protest-related channels as a justification for her arrest (Belsat 2020c).

Belsat TV coverage of Telegram also included several in-depth pieces, including opinion columns and interviews. On 17 August, the outlet also interviewed NEXTA_Live’s editor Putsila, who, inter alia, suggested that the reason people flocked to Telegram channels was that other sources of information were either co-opted by the state or inaccessible due to censorship (Belsat 2020b). An 11 August opinion piece comparing protests in Belarus and Ukraine discussed the role of internet and social media in Belarus and noted the central role of Telegram channels in disseminating information despite state censorship of mainstream media (Pogozhelskiy 2020). On 5 September, a political consultant opined that Telegram channels “provoke heightened activity among the public” and said more internet shutdowns should be expected (Feduta 2020).

State-run or state-friendly media outlets, Belta and Komsomolskaya Pravda Belarus (KP), mostly mentioned Telegram in the context of citing state Telegram channels (e.g. Lukashenka’s press service or the spokesperson for the Interior Ministry of Belarus) as sources of information. However, both outlets also framed protest-related Telegram channels as “run by agents of the West interested in fomenting conflict” or as “destructive,” assigning them at least partial agency.

In this regard, KP was somewhat more balanced, reporting on blocked independent media outlets who now had to rely on Telegram channels and mentioning that Telegram channels were directing protesters and organizing rallies. A 15 August analysis piece by a local KP correspondent compared Belarus protests to Ukraine’s Euromaidan and said the former were really grassroots and were directed by “Telegram channels” and not “oligarchs” (Komsomolskaya Pravda Belarus 2020).

Belta extensively cited state-run Telegram channels as sources, and framed these in entirely neutral terms. In contrast, protest-related Telegram channels were framed as dangerous spaces full of provocations where citizens could find instructions for making Molotov cocktails or send anonymous threats to civil servants and police officers and leak their personal data. Belta’s coverage implied that Telegram channels and their moderators instigated more violent and aggressive behavior among protesters, who were incited to block roads during rush hour or offered money to participate in protests.

Belta also ran several in-depth analysis or opinion pieces referencing Telegram. In a 28 August opinion piece, a Belarusian military expert said that Belarus protests showed “signs of hybrid warfare” and that “destructive Telegram channels” were acting “based on the textbooks of color revolutions” (Belta 2020c). In another op-ed published the same day, another state-friendly political expert said Belarus was seeing “a four-stage hybrid warfare scenario” and opined that Belarus was in the first stage of a social media-driven revolution: “Whereas before, in Egypt, in Tunisia something similar was done through Facebook, we have a clear Telegram-revolution” (Belta 2020f). This was echoed by a Ukrainian political scientist in a 27 August piece, who agreed that Telegram channels were run by “foreign agents” from abroad and that Telegram was instigating a new Twitter revolution, akin to the Arab Spring (Belta 2020e). In a 10 September piece, a Russian expert posited that the Belarus protest was “a network protest” without leaders run “by Telegram channels” (Belta 2020d). The frame of Telegram channels as puppeteers manipulating Belarusian citizens from abroad was pervasive in Belta’s coverage.

State perception

The key frame of Telegram and Telegram channels (sometimes named, often nameless) as “destructive” actors manipulating citizens or as “foreign agents” working to destabilize the situation in Belarus was dominant in public statements of Belarusian state officials. This can be interpreted as evidence of the state assigning certain Telegram channels (and, by proxy, Telegram) a measure of political actorness in the context of the protests. On the other hand, such demonization presents a convenient, though abstract, adversary for the Belarusian state to blame for the growing unrest.

Lukashenka’s pronouncements on Telegram’s role in the protests featured most prominently in the coverage of both state and independent media. A Belta piece on 4 August 2020 quoted Lukashenka as saying that Belarus was embroiled in “information warfare” in a world rife with “hybrid wars and color revolutions” and that social media and Telegram channels in particular, “have been flooded with nonsense, dirt, and shameless lies” (Belta 2020g). A report by NavinyBy from 9 September cited Lukashenka remarking to Russian media that it was impossible to fight Telegram channels and that Belarus and Russia were facing the same struggle. He also complained that Telegram and other social media had “scrambled the youths’ brains and brought disruption to ‘quiet and calm Belarus’” (NavinyBy 2020b). In a 10 September Belta interview, claiming Belarus was undergoing a coup attempt, Lukashenka cited both external influence and local organizers “executing the politics of these Telegram channels” to organize anti-government protests (Belta 2020h); in another 24 September comment to Belta, he implied that the Telegram channels that had leaked information on his secret inauguration were managed from abroad by the CIA, German, and Polish intelligence.

Other state officials echoed Lukashenka’s framing: a Belarusian MP announced on 8 September that protests were coordinated through Telegram chats, including NEXTA_Live, from abroad (Poland), as reported by Belsat TV (Belsat 2020d); the Deputy Prime Minister was quoted by Belta on 27 August as saying that Telegram and other messengers had significant influence on citizens by “disseminating fakes,” inciting discontent, and “manipulating the public consciousness” (Belta 2020a). Belta also reported on the Interior Ministry spokesperson warning on 24 September not to react to calls from “destructive” Telegram channels that provoke people to violate the law, while another Belta report on 27 September featured the deputy Health Minister arguing that political events “orchestrated by Telegram channels” had undermined public perception of Belarus’ efforts to fight COVID-19 (Belta 2020b).

The state’s perceptions of Telegram as an actor implicated in the growing discontent were reflected in repressive actions against the platform. As Telegram channels gained prominence, they and their administrators became targets for repression amidst a wider crackdown on independent media, journalists, and human rights organizations. Up to 24 July 2021, 107 channels were designated as extremist, including the “97% Minsk” channel analyzed in this study (A Country to Live In 2021). Sharing any content from these sources meant facing fines, community service, or administrative arrest (Article 19.11 of the Belarusian Code of Administrative Violations). Several Telegram channel administrators had been detained throughout 2020, while Raman Pratasevich and Stsyapan Putsila, the administrators of the NEXTA_Live channel, were accused of organizing mass riots and placed on the KGB’s “List of organizations and individuals involved in terrorist activities” in November 2020 (RFE/RL Belarus Service 2020). Pratasevich was arrested by Belarusian authorities on 23 May 2021 after his plane, traveling from Athens to Vilnius, had been diverted to Minsk on false pretenses (RFE/RL Belarus Service 2021).

User perceptions

The state’s concerns about Telegram’s involvement in the unrest were echoed by user preoccupation with state surveillance and app accessibility. As new users joined the group ahead of the presidential elections, chat conversations demonstrate that concerns about anonymity on Telegram were widespread. New members inquired whether their names or telephone numbers were visible, while less privacy-aware users were reminded not to use images that could be traced back to their person via a reverse image search. Interestingly, while such questions make up a significant proportion of posts at the beginning of our sample, only one such mention was found for the period 9–10 August. This suggests that either most users had already joined the group at an earlier time, or that, as internet shutdowns took effect, concerns about anonymity were overtaken by more immediate concerns about access to information.

Chat group discussions on 9–10 August provide insight into how Belarusians sought to negotiate the internet shutdown and assessed their options for (re)gaining access to information. Various VPN services (e.g. Tunnel Bear, Psiphon) and Telegram proxies, other social media platforms and messengers (VK, YouTube, Twitter, Instagram, Viber, Telegram, WhatsApp, Zello), and websites (Tut.by, Onliner.by, TVRain.ru, Yandex.ru) were discussed. Whether or not a particular service or platform worked at all soon became the central concern. For example, users reported difficulties accessing Belsat TV’s livestream on YouTube. Group admins played an important role by continuously sharing links for downloading VPNs that reportedly still worked, along with the NEXTA_Live channel sharing links to new proxies. But those who had not installed VPN before the internet shutdown encountered difficulties as app stores became inaccessible and torrent download speeds dropped, often leaving Telegram-specific proxies as the only remaining option.

While on 9 August a wide range of VPNs and platforms was being discussed and recommended, by 10 August recommendations leaned heavily towards using Psiphon and/or Telegram-specific proxies, with the latter appearing to have been most resilient (confirming the platform’s actions to technically facilitate use were successful). The analyzed chat conversations therefore suggest that people did not necessarily want to use (only) Telegram; rather, they sought to access sources of information they trusted, some of which (in particular, NEXTA_Live) were on Telegram. On 9 August, we also identified many questions about and reports of problems with submitting a picture of one’s ballot to the grassroots election monitoring initiative Golos. Here, the Telegram bot enabling submissions again appears to have been the most stable, compared to the Golos website and its bot on Viber.

On 9 August, as protesters gathered on the streets, participants were sceptical about claims made in the chat and asked to see proof (pictures, video), which was hampered by limited connectivity. Information sourced from NEXTA_Live was not questioned, which aligns with survey data indicating that “Telegram [was] regarded as the most trusted news source and as such [was] used by many of those who protested in person (85% [of survey participants, author])” (Herasimenka et al. 2020). Some also warned peers the regime was monitoring the chat and potentially seeking to manipulate it. Accusations of trolling behavior intensified on 10 August, especially concerning messages about the use and preparation of Molotov cocktails (these instances were also reflected in our analysis of media coverage above). In the evening, as users shared addresses for where to find safe shelter and entrance codes for residential buildings, the chat was “flooded” by expressions of support from abroad (“Rossiya s vami!”; “Riga s vami!”). Users asked for the spamming to stop and for admins to intervene. The chat was temporarily suspended several times, and after some time the flood of support messages abated, suggesting the admins had successfully fought off the attack.

When speaking about content moderation, chat participants mostly addressed the group admins: to ask them to share information about VPNs, to block users who they believed to be provocateurs, or to temporarily suspend the chat when overrun with trolling. We did not identify references to moderation decisions by Telegram itself within the 9–10 August timeframe. Similarly, direct references to Pavel Durov (N = 26 between 22 June and 23 August 2020) were positive. Users expressed their gratitude for the fact that Telegram worked while nothing else did, or claimed that Durov did not collaborate with the security services but, instead, supported the Belarusian people. Statements by Durov were shared in the group, including, on 25 June, Durov’s assertion that achieving full anonymity online was impossible (the user who shared it advised the chat to use Conversation and Protonmail instead).

Overall, chat group participants’ use and perceptions of the platform were shaped by pragmatic concerns about remaining connected and having access to information, but also by their perception of risk with regard to, for instance, anonymity and security. However, they also exhibited a level of unquestionable trust in Telegram channels as a source of information, while also remaining alert to possibilities of hostile actors infiltrating the chats. Attitudes toward the platform and its figurehead, Durov, remained predominantly positive throughout the period we analyzed, as evident from both indirect perceptions of the platform’s actorness through specific channels and groups, and direct mentions of the platform’s availability, circumvention tools, or its founder’s statements.

Conclusion

In this article, we have argued for the need to approach digital platforms as actors in contentious politics, rather than seeing them as mere tools for political participation. Expanding upon recent work on platform governance and platform power and adapting it to the particularities of contentious politics, we have proposed a conceptual framework for analyzing platform actorness and applied it to the case of Telegram in the 2020 Belarus protests. Analyzing Telegram’s enacted, performed, and perceived actorness, our research finds that Telegram’s performance and practices resonated with Belarusian citizens and led them to form affective connections to the platform, perceiving Telegram as an ally in their struggle against pervasive repressions and digital censorship. At the same time, the Belarusian state used Telegram’s aversion to censorship and preference for limited content moderation as an opportunity to intervene in contentious politics by co-opting grassroots civic approaches and mimicking manipulative efforts of other authoritarian regimes, showing a propensity for authoritarian learning, while also using Telegram as a convenient scapegoat to blame for the discontent.

As Belarusian users weighed pragmatic considerations and a desire to build and sustain communities of action against their perceptions of risk, the platform’s reputation, as well as the reputation of particular channel and group administrators, played an important role in decisions about using Telegram. The platform’s actions were key to enabling access and community-building (technical facilitation of use), but they also helped build a sense of allegiance necessary for sustained use even after internet shutdowns ended (changes in platform design, including releasing white-red-white themed stickers). At the same time, the platform’s lack of content moderation made protesters’ communications, and by extension their material actions, vulnerable to manipulation and provocations that have also served as pretext for prosecutions.

The case of Telegram demonstrates that digital platforms indeed have agency in ways that are relevant to contentious politics. As affective relations are known to play an important role in how communities of action form and consolidate, we argue that such affective relations are also formed vis-à-vis platforms – a dimension that requires further research. For users and researchers alike, gauging platforms’ intentions is complicated by two factors. First, in their performance, platforms are “unreliable narrators”: due to their business model (whether based on venture funding, user data sales, or advertising income), most platforms favor growth of user numbers and engagement, which comes to shape the rest of their practices. Yet, in their public relations, platforms often make strategic use of particular civic imaginaries to present a partial, yet appealing image. For example, Telegram’s corporate communications strategy reinforces personalization of corporate values and identity in the figure of Durov, obscuring the more mundane realities of a business vying for survival in a global capitalist marketplace. Second, users’ perceptions of platforms and particular persons and groups active on those platforms risk becoming conflated: the reputation of a particular channel may easily be projected onto the platform, or vice versa. Both factors may lead to incomplete or incorrect assessments of risk and reliability, with severe consequences at politically contentious times.

Though our empirical work investigating Telegram’s actorness is made more challenging by the difficulty of assessing citizens’ and state perceptions through extant data, it extends our understanding of the complicated role of social media platforms in contentious politics in the post-Soviet space. Our findings are applicable to the region where scholarship finds considerable evidence of authoritarian learning, as they help understand not only how citizens perceive and use digital platforms, but also how post-Soviet autocracies may co-opt and exploit these perceptions.

Our research contributes to understanding how platforms’ actions and the (lack of) debate around their work in authoritarian states (especially those considered “low priority”) may perpetuate otherwise discredited strands of thought about the role of platforms in democracy promotion. While our analysis builds on a case of contentious politics in an autocratic regime, our conceptual framework for platform actorness can equally elucidate the triangle of platform-state-citizen relations in other regime types and therefore has global applicability. Given that digital platforms are increasingly central to political processes, from protests to elections, in democracies as well as authoritarian states, our framework can serve as a useful heuristic for investigating the differing degrees of actual and perceived responsible behavior with respect to different target audiences and markets. Future research in this arena would extend our understanding of platform actorness in various political contexts and add to the growing body of literature examining platform power, transparency, and accountability.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Content moderation refers to the organized practice of screening user-generated content on social media platforms to assure they are in line with platform guidelines and applicable laws.

2. End-to-end encryption ensures only the sender and receiver of a message can read it, while those intercepting the data see it as a string of meaningless characters.

3. For more on the challenges and risks of platform content moderation, see Gillespie (2018) and Gorwa, Binns, and Katzenbach (2020).

4. Here, domestic signifies by and for a particular national audience, rather than produced on the territory of the respective state.

5. The authors gratefully acknowledge the invaluable assistance of Dr. Samuel Greene, King’s College London, in obtaining access to the Telegram group chats dataset.