The quest for exposure: How pro-Palestine activists navigate algorithmic visibility and content moderation on TikTok

Abstract

While political content, particularly from warzones, is frequently subject to moderation and censorship on social media platforms, pro-Palestinian creators face uniquely constrained conditions when speaking up for their cause. Often physically confined, they use digital platforms to connect with diasporas, share everyday struggles, and amplify narratives to a global audience. On these platforms, they are both hyper-visible and hyper-censored, challenged by opaque moderation systems, anti-Arab biases, and narrative erasure driven by platforms’ monetary interests and political pressures. This study examines how pro-Palestinian activists on TikTok navigate these challenges through qualitative interviews with 13 content creators. Participants describe experiences of visibility moderation and content removal, often under vague or inconsistent enforcement. These accounts align with the concept of platform gaslighting, in which unclear explanations and ineffective appeals foster uncertainty and mistrust in platform governance. Yet our findings reveal that, despite these pressures, the well-connected creators we interviewed continue to develop creative strategies to maintain visibility and avoid suppression. Through collective action, they expose misleading community guidelines, contest platform gaslighting, and resist silencing, reflecting a broader struggle over freedom of expression in algorithmically governed spaces.

Keywords

TikTok activism content moderation platform gaslighting Palestine shadowbanning

Introduction

In the rapidly evolving landscape of digital communication, social media platforms have become integral to shaping public discourse, influencing societal narratives, and fostering political engagement (Bennett and Segerberg, 2012; Lievrouw, 2023: 1952–1969; Shumate and Lipp, 2008). A prime example is the Chinese short video app TikTok, which has gained immense popularity in the last couple of years. TikTok is especially popular among the younger demographic. It has over a billion active users, of which 38.5% are between 18 and 24 years old and 32% are aged between 25 and 34 in the year 2022 (Omnicore, 2023). The platform's design for viral communication makes it a powerful tool for online political activism (Cervi and Marín-Lladó, 2021; Hautea et al., 2021).

This potential has also been recognized by pro-Palestinian content creators, who use TikTok to contest narrative erasure and make their political realities visible. In this study we focus on pro-Palestinian activism on TikTok. Drawing on interviews with thirteen content creators, we offer a user-centered account of how activists interpret and navigate algorithmic moderation amid the ongoing war in Gaza. In doing so, we advance debates on algorithmic imaginaries, algorithmic gossip, and platform gaslighting within a geopolitically asymmetrical context.

TikTok employs a combination of automated systems, human moderators, and recommendation algorithms to determine what content appears on its default feed, the For You Page (FYP), and in other content streams across the platform (TikTok Help Center, 2026). Understanding how these systems operate, which content is recommended, suppressed, or deleted, is crucial for content creators, especially for political activists aiming to spread their messages and mobilize people (Abbas et al., 2022; Cervi and Marín-Lladó, 2021; Klug et al., 2021; Le Compte and Klug, 2021; Zeng and Kaye, 2022). While moderation is essential to ensure the platform adheres to community guidelines and to prevent the spread of harmful or inappropriate content such as disinformation and hate speech, its implementation often disproportionately affects activists and marginalized groups (Gorwa et al., 2020; Biddle et al., 2020), who may find their content flagged or removed for reasons that are unclear. This can be particularly problematic when attempting to disseminate politically sensitive or dissenting viewpoints (Gillespie, 2022; Duffy Meisner, 2023).

Table 1.

Overview of participants with follower counts and total likes on TikTok.

Participant no.	Followers	Likes
1	19,620	1.5 mio
2	158,159	6.7 mio
3	63,812	1.8 mio
4	111,455	2.5 mio
5	102,334	11.2 mio
6	8745	84,519
7	125,680	7.7 mio
8	41,208	1.3 mio
9	19,429	2.3 mio
10	263,762	2.1 mio
11	123,061	6.9 mio
12	92,600	4.2 mio
13	11,651	350,534

The lack of transparency regarding TikTok's recommendation and moderation decision-making processes (Gillespie, 2022; Zeng and Kaye, 2022) has led users to develop informal assumptions about how the algorithm operates and to craft strategies aimed at boosting visibility or avoiding invisibility (Bucher, 2017; Duffy and Meisner, 2023; Le Compte and Klug, 2021; Savolainen, 2022; Steen et al., 2023). Bucher (2017) conceptualized these user-formed assumptions as algorithmic imaginaries, a term that has since been expanded through related concepts such as algorithmic gossip (Bishop, 2019) and algorithmic folklore (Karizat et al., 2021), which all describe how users make sense of systems that deeply affect their everyday lives but remain opaque. In contexts of politically sensitive content, Divon et al. (2025) and Are (2024b) introduce the notion of platform gaslighting, focusing on how these systems are experienced by the users, who are led to believe that moderation failures are individual errors rather than systemic patterns, contributing to uncertainty, silencing, and a loss of trust in platform governance.

An ongoing political event where social media and TikTok have played a significant role is the militant occupation of Gaza and the conflict between Israel and Palestine, which the media have described as “TikTok intifada” in 2021 (Divon, 2022) or a new War Zone (Li and Prasad 2018: 505). The Hamas attack on Israeli citizens on 7 October 2023 has reignited and further escalated the conflict. Israel's military response to the terrorist attack led to the destruction of Gaza with widespread casualties and displacement, which a 2025 UN Commission found to constitute genocide (Office of the United Nations High Commissioner for Human Rights, 2025).

As pro-Palestinian content creators seek to rally support online, they frequently report that their content is being suppressed (Shankar et al., 2023). While political content, particularly from warzones, is frequently subject to moderation and censorship on social media platforms (Banchik, 2021; Elmimouni et al., 2024), pro-Palestinian creators find themselves in a uniquely constrained position when advocating for their cause on social media platforms that are heavily influenced by Western platform norms and moderation standards (Roberts, 2019: 33–72). Often physically confined, they have turned to digital platforms to connect with diasporas, share their everyday struggles, and amplify their narratives to a global audience (Cervi and Divon, 2023).

On these platforms pro-Palestinian activists are both hyper-visible and hyper-censored. They must contend with opaque moderation systems, anti-Arab biases, and narrative erasure driven by platforms’ monetary interests and political pressure (Elmimouni et al., 2024; Roberts, 2019: 33–72; York, 2022: 73–100). In contrast to other users, their visibility challenges are not merely algorithmic but deeply politicized, shaped by decades of occupation, digital repression, and geopolitical silencing. They exemplify how non-Western, marginalized, and subaltern activists navigate the contradictory terrain of social media, platforms that promise visibility and voice while simultaneously enforcing norms that suppress their narratives and existence (Divon et al., 2025; Cervi and Marín-Lladó, 2022).

In light of TikTok's significant role in political engagement and activism among youth, this study investigates how pro-Palestinian activists experience content moderation on TikTok. We focus on undisclosed algorithmic visibility moderation (e.g., shadowbanning) examining the strategies activists employ in response. While existing research has examined activism on TikTok, much of it has focused on content-level analysis or policy review. This study expands the field by offering a timely, interview-based account of how activists themselves experience, interpret, and respond to such moderation. This approach not only deepens our understanding of the algorithmic imaginary but also underscores the importance of more recent concepts such as platform gaslighting, highlighting the emotional, strategic, and political labor and stress involved in navigating visibility under conditions of asymmetrical moderation and geopolitical marginalization.

We focus on the following two research questions:

RQ1: How do pro-Palestinian activists experience algorithmic visibility moderation on TikTok in their quest for algorithmic exposure?

RQ2: What strategies do pro-Palestinian activists use to navigate algorithmic visibility moderation on TikTok?

Literature review

Social media and political activism

Social media platforms have revolutionized political activism by providing new avenues for collective action, political mobilization, and information dissemination (Lievrouw, 2023). Movements such as the climate justice protests during COP15 in Copenhagen, the Arab Spring, and Black Lives Matter (Askanius and Uldam, 2011; Cox, 2017; Eltantawy and Wiest, 2011) demonstrate how digital technologies enable activists to challenge dominant media structures, bypass traditional media gatekeeping, circumvent censorship, and enable direct, unmediated communication with citizens (Askanius and Uldam, 2011; Garrett, 2006). Each of these moments also marks an intersection where activism and platform infrastructures reshaped one another; they established both the affordances and the vulnerabilities of digital activism that today shape Palestinian online resistance.

Social media platforms facilitate what Bennett and Segerberg (2012) call connective action, a decentralized, crowd-enabled mode of organizing that is amplified by social media affordances and platform logic (Abbas et al., 2022; Tufekci, 2017). The platforms focus on photos and videos also offer new possibilities to show everyday realities in war, occupation, and resistance, evolving earlier digital tools into highly affective, networked testimonial accounts that circulate widely on social media (Askanius, 2013). As Heřmanová et al. (2025) note, “the algorithm loves the war,” often amplifying sensational or graphic conflict imagery while muting contextual or critical perspectives, a pattern acutely visible in how Palestinian realities are mediated online.

Yet this potential has always been marked by ambivalence. Social media are not neutral infrastructures; they operate under commercial platform governance (Uldam, 2018), and activists must constantly navigate opaque algorithms, moderation regimes, and surveillance mechanisms that shape not just what can be said but also whether it can be seen (Gillespie, 2022). This is especially evident in contexts of conflict, where platform policies intersect with geopolitical narratives. Although often celebrated as a moment of digital empowerment, the Arab Spring has since been re-evaluated as a case of platform disillusionment, algorithmic depoliticization, and state surveillance. Platforms initially praised for enabling protest were later seen as complicit in depoliticizing activism and facilitating state-led surveillance and repression (Tufekci, 2017; Zayani and Khalil, 2024: 88–101).

The same is true for Palestinian activism on social media, which dates back to the early 2000s with the expansion of internet access in the occupied territories (Monshipouri and Prompichai, 2017). While social media has offered new opportunities to overcome physical separation and connect with the diaspora in a highly fragmented media environment, the Israeli occupation has consistently extended into the digital sphere in what Tawil-Souri (2012) terms “digital occupation,” characterized by Israeli control over telecommunications infrastructure and legal frameworks (Mahlouly and Erhaim, 2023: 237–250). From early experiences on Facebook and YouTube, where takedowns, shadowbans, and biased moderation were commonplace, these constraints have shaped a decades-long struggle over digital visibility for Palestinian voices (Mahlouly and Erhaim, 2023; Tawil-Souri, 2012). Elmimouni et al. (2024) show that such dynamics persist on newer platforms: During the Sheikh Jarrah crisis, when a long-standing land dispute in East Jerusalem escalated in 2021, content moderation involved systemic over-enforcement and keyword-based suppression. Recent publications confirm that these practices continue: 7amleh's Hashtag Palestine 2024 report documents widespread and ongoing moderation of pro-Palestinian content across platforms, including TikTok, through takedowns, shadowbans, and restrictions on key political hashtags (7 amleh – The Arab Center for the Advancement of Social Media, 2025).

TikTok's recommendation algorithm and content moderation

As a newer platform, TikTok combines established social media logics with unique affordances that support collective and connective action (Abbas et al., 2022). Its large, youth-driven user base (Omnicore, 2023) and viral-oriented recommendation system (Bandy and Diakopoulos, 2020; Klug et al., 2021) make it a powerful tool for political mobilization and activist visibility. Key affordances of TikTok include the low-entry barrier for content creation, features like duets, stitches, and challenges that encourage participation by allowing users to collaboratively respond to, reinterpret, and remix content (Cervi and Marín-Lladó, 2022; Cervi and Divon, 2023), and most importantly, the recommendation algorithm that governs visibility and virality (Klug et al., 2021; Guinaudeau et al., 2020; Zeng and Kaye, 2022).

Unlike social media platforms where visibility is often tethered to follower networks (e.g., Instagram or Facebook), TikTok's default feed, the FYP, is powered entirely by algorithmic recommendations based on user interaction and engagement patterns such as watch time, likes, comments, and shares (Klug et al., 2021; Guinaudeau et al., 2020). This system affords any video, regardless of the creator's follower count, the potential to go viral if it performs well with an initial sample audience (Abidin, 2020; Matsakis, 2020).

While TikTok's algorithm offers opportunity, it also serves as a central mechanism for content moderation (Zeng and Kaye, 2022). TikTok states that content moderation is essential for ensuring user safety and combating harmful content such as hate speech and disinformation (TikTok Help Center, 2025), but scholars argue that these practices are driven by commercial and brand-protection interests (Gillespie, 2018; Roberts, 2019; Are and Paasonen, 2021). Like other platforms, TikTok categorizes and ranks content using a combination of engagement signals and automated content analysis, including computer vision and natural language processing (NLP) techniques to analyze video content, hashtags, and descriptions (Scanlon, 2020; TikTok Help Center, 2023). This results in a dual function: The same algorithm that amplifies some content can suppress other forms, particularly those deemed controversial, political, or in violation of platform norms (Zeng and Kaye, 2022; Gillespie, 2022; Brown, 2021; Bacchi and Reuters, 2020).

Zeng and Kaye (2022) refer to this as algorithmic visibility moderation: the use of recommender systems not only to promote content but also to quietly limit or eliminate its visibility without formal removal. Algorithmic visibility moderation is now an important part of the toolkit of platform governance, alongside more explicit forms of moderation such as demonetization, content removal, or deplatforming (Gillespie, 2022; Are, 2023). While platforms like TikTok do not publicly acknowledge algorithmic visibility moderation as an official policy, users report significant drops in reach or engagement following posts on politically sensitive topics, with no explanation or appeal process (Are, 2024a; Nicholas, 2022). Such undocumented moderation becomes part of what Bucher (2017) calls the algorithmic imaginary: the ways users imagine and interpret opaque systems based on their lived experience.

In such environments, where rules are unclear or inconsistently enforced, creators frequently share algorithmic gossip, informal, and collective theories about how platforms work, as a means of risk management (Bishop, 2019). Also conceptualized as folk theories (Karizat et al., 2021), these knowledge practices are critical, especially for marginalized users whose content is more likely to be suppressed. Such theorizing reflects the reality that platform moderation is “strategically opaque” (Gillespie, 2022), designed to maintain control and avoid accountability without overt censorship. This systemic discursive strategy, where platforms deny or deflect users’ experiences of moderation harm, often implying that users are mistaken or imagining the problem, can contribute to the aforementioned platform gaslighting (Divon et al., 2025). Are (2024a) demonstrates how queer users, sex educators, and creators operating in repressive regimes navigate such moderation through practices like self-censorship, platform circumvention, and the development of counter-discourses. Moreover, this system of algorithmic gatekeeping is not neutral. Studies have shown that moderation practices disproportionately impact users from marginalized communities, often reflecting geopolitical, racial, or ideological biases embedded in platform governance (Biddle et al., 2020; Brown, 2021; Bacchi and Reuters, 2020; Köver and Reuter, 2019). As Elmimouni et al. (2024) demonstrate in the context of the Sheikh Jarrah crisis, moderation was not only about removing harmful content but also about the systematic erasure of narrative, for example, through the suppression of keywords like “Palestine” or the removal of entire stories without cause. This unfairness is further exacerbated by concepts such as algorithmic popularity bias (Abdollahpouri et al., 2020). Content that already receives high engagement is more likely to be surfaced again, while content suppressed early in its distribution cycle may never recover.

Activists’ strategies for navigating algorithmic (in)visibility

In response to opaque algorithmic governance and perceived suppression, activists on TikTok have developed a range of strategies aimed at maintaining visibility and circumventing moderation while preserving the integrity of their political message. These strategies form what Cotter (2019) calls the “visibility game,” wherein users deliberately tailor their behavior to align with what they believe the algorithm rewards, a mix of creativity, compliance, and constant calibration. A foundational tactic involves aligning political content with TikTok's viral formats and participatory culture. TikTok's fundamental design as a “memetic platform,” as conceptualized by Zulli and Zulli (2022), underscores that visibility and virality on the platform are inherently driven by modular remixing, imitation, and memetic replication. Consequently, activists leverage trends, sounds, memes, and challenges to embed activist messages within platform-native styles of entertainment (Hautea et al., 2021).

Scholars describe this as politainment, the blending of political messaging with entertainment logics to increase engagement and reach (Cervi et al., 2023; Gonzalo et al., 2021; Nieland, 2008). This form of what Cervi and Marín-Lladó (2022) and Cervi and Divon (2023) term playful activism is not merely aesthetic. It reflects a strategic use of platform affordances to make dissent visible under constraints of moderation. Their work shows how young Palestinian TikTok users employ emoji-based vernaculars, visual metaphors, and performative embodiment to express solidarity and resistance without triggering takedowns. A strategy that has been further conceptualized by Divon and Eriksson Krutrök (2024) as the “platformed body,” wherein users offer their bodies to the algorithm as both medium and message, amplifying trauma, dissent, and visibility. These examples all emerge from active war zones, where activists must navigate the tension between wartime realities, platform affordances, and content moderation. As Heřmanová et al. (2025) note, such contexts produce an ambivalent visibility: While moderation frequently targets violent imagery (Banchik, 2021; Elmimouni et al., 2024), TikTok's algorithm often amplifies war content, once it passes initial moderation barriers.

Another widely reported strategy is the use of coded language or algospeak, deliberately substituting sensitive terms with phonetically similar or semantically adjacent ones (e.g., “Pal3stine”) (Steen et al., 2023). These tactics allow activists to evade automated moderation filters while signaling meaning to an in-group audience. They hint at the constant precarity of activist expression online, and the labor involved in staying visible (Duffy & Meisner, 2023). Activists do not only think about how they share and package content but also what content they do or do not share. Faced with repeated content moderation, some users report avoiding certain topics, images, or hashtags altogether (Duffy and Meisner, 2023). Such self-censorship can be seen as a chilling effect, wherein the threat of moderation leads users to withhold legitimate expression (Penney, 2016; Büchi et al., 2022). For marginalized activists, this is not a theoretical concern but a reality that shapes what can be said, and how. Self-censorship underscores that these tactics are not merely cosmetic; they can fundamentally alter the message itself, constraining political expression (Duffy and Meisner, 2023).

Methodology

Recruiting and data collection

Our study draws from critical platform studies and digital ethnography, focusing on how users experience and interpret opaque algorithmic systems. We conducted semi-structured interviews with pro-Palestinian activists for an in-depth examination of participants’ perspectives and experiences with the algorithmic mechanisms of TikTok. This approach is crucial to understanding the motivations behind specific content creation strategies and the contextual factors influencing them. Our aim was to produce interpretive and experiential knowledge, foregrounding participants’ subjective understandings, algorithmic imaginaries (Bucher, 2017), and resistance practices. The flexibility of semi-structured interviews allows for detailed exploration of topics, revealing unexpected patterns and diverse viewpoints. Additionally, this approach encourages active participation, leading to richer data collection.

We started recruiting by identifying potential creators from the second author's personal TikTok following list and “For You” page and searched for hashtags related to the pro-Palestinian movement to locate suitable creators. We defined a pro-Palestinian stance through markers such as political hashtags (e.g., #freepalestine), visual markers such as Palestinian flags, keffiyehs, and explicit content supporting Palestinian rights. We define activism as the intentional use of TikTok to raise awareness and mobilize support. Our focus is on overt (algorithmic) resistance by creators explicitly advocating for Palestinian rights. Inclusion required at least 10 politically themed videos, with a minimum of 5 explicitly advocating for Palestinian rights within the past year. We set these thresholds to ensure that participants had a sustained and recent record of political posting rather than isolated or episodic activity. Pilot screening suggested that accounts below these thresholds lacked sufficient material to support robust, comparative analysis of strategies and experiences. We acknowledge that algorithmic resistance is often subtle, and this excludes many ambiguous cases we encountered, where intent was unclear or context lacking, such as using filters or sounds connected to the cause without clear political messaging. Excluding these cases carries important limitations, as it may have left out smaller creators operating at the margins of visibility, who might be more vulnerable to silencing and algorithmic gaslighting.

After confirming suitability based on the participant's prerequisites (see Participants), we made contact via Instagram direct messages or e-mail (if provided in their TikTok bio). We chose e-mail and Instagram direct messages as to initiate a conversation with a person on TikTok, one must follow each other. This was challenging due to our lack of posted content and followers. We created a Google Docs document including all the information of the study (Appendix 7.1). When contacting the potential participants on Instagram, we included a link to the Google Doc document, because Instagram limits users’ direct message requests to a certain number of characters. Despite observing and engaging with the TikTok activist community, we had not actively posted political content or established connections within these circles. This may have caused caution among community members when interacting with us. Furthermore, asking for a 30 min interview without offering any compensation can be a demanding request. This potentially affected the motivation for individuals to participate.

We collected data for this study via qualitative in-depth semi-structured interviews conducted between 18 March and 21 April 2024, with n = 13 English-speaking pro-Palestinian TikTok political activists. These took place via Zoom and lasted between 20 and 40 min. We designed the questions (Appendix 7.3) particularly to better understand participants’ general motivation for specific content creation strategies. We kept the interviews concise to avoid overburdening participants from a community already under pressure. A narrow focus on our RQs allowed us to gather a broad range of perspectives within a clearly defined research scope. Despite the relatively short interviews, many participants gave similar answers, sharing similar experiences which suggests that we reached a good level of saturation.

Participants

We established several prerequisites for the participants of this study. Firstly, the participants had to be active content creators on TikTok. They had to have posted at least 10 videos related to political issues on TikTok in the last year. Furthermore, they had to create political content advocating for Palestinian rights under the Israeli occupation. A minimum of five videos related to this topic must have been posted in the last year. We also tried to identify creators willing and motivated to discuss their experiences with the TikTok recommendation algorithm, especially regarding content visibility and moderation. For recruiting, we did not consider the participant's follower count, as it is not crucial for content visibility on TikTok, unlike other platforms (Bandy and Diakopoulos, 2020). The participants had relatively large follower counts ranging from 8000 to 260,000, with an average of 87,000 followers. Regarding engagement metrics, the range for participants’ most liked or viral videos spanned from 84,000 to 11 million likes, with an average of 3.7 million likes (see Table 1). The participants are a diverse group in terms of age, background, and content focus. Some are Jewish, some Muslim (we did not explicitly ask about religious identity or location due to privacy concerns), and some are part of the Palestinian diaspora. Their TikTok content ranges from educational videos on history and politics to entertainment-based content that occasionally engages with political issues such as the conflict in Palestine. Some identify as activists both online and offline, while others use their platforms to raise awareness without identifying primarily as political creators. Most are based in the USA, followed by the UK, with a few from other locations. A few interviewees chose to remain fully anonymous.

Data analysis

First, we transcribed the audio-recordings of every interview. Next, we used the open-source qualitative data analysis software QualCoder and coded the audio-recordings in two steps. Our coding process was primarily inductive, following the principles of Grounded Theory (Charmaz, 2006). While guided by our research questions, initial codes emerged organically from the data. These were then consolidated into broader themes through iterative comparison and collaborative discussion. Initially, we coded the transcripts for themes and gave them a name and a description, including the importance of each theme for the research questions. For example, one theme was the uncertainty creators expressed about why their videos had been taken down, which we grouped under a broader category of experiences with content moderation. Similarly, themes around perceived unfairness in appeal processes or algorithmic traction were categorized as part of the larger challenges creators face. Another set of themes focused on the corresponding resistance strategies, such as adjusting language or metadata to avoid being flagged or encouraging engagement to boost algorithmic visibility. Those themes grouped into larger categories formed the basis of our data analysis.

When conducting the interviews, we were aware of our own backgrounds, including biographical, educational, sociocultural, gender, racial, and sexual identity. As white researchers from a university in the Global North, we recognize that our positionalities may have influenced both who chose to participate and how openly they shared their experiences. We may not have reached activists who are more mistrustful of academic institutions or who share more radical content, and some participants may have hesitated to speak fully due to perceived lack of safety. Although we did not observe any explicit signs of discomfort or interviewer effect, we acknowledge that such influences may be subtle and not always noticeable in the moment (West and Blom, 2016). Further, we acknowledge that language-related biases cannot be fully ruled out. The interviews were conducted in English, even though English was neither the first language for some of the participants nor for us.

Ethical considerations

Ethical considerations guide every aspect of this study, ensuring participants’ welfare, confidentiality, and the handling of their data responsibly. This study aligns with the ethical principles outlined by the Association of Internet Researchers (AoIR, 2019) to ensure research integrity and participant protection in online studies. Regarding transparency, participants have been thoroughly informed before interviews about the study's purpose, risks, and methods to ensure their clear understanding and voluntary participation. They had the possibility and the right to seek clarification at any stage, supporting their autonomy and participation without coercion. Anonymity and confidentiality were paramount to protect them from potential repercussions related to their political activism on TikTok, ensuring a secure environment for open discussions. Furthermore, interview questions have been reviewed for cultural sensitivity, aiming to respect diverse political views and avoid offending or misrepresenting participants. The collected data had been used solely for the purpose of this study, maintaining integrity and trust in handling participants’ information. Finally, this research aims to contribute positively by enhancing the understanding of TikTok's political activists’ experiences with algorithmic visibility. This benefit was carefully weighed against any potential risks or discomfort that participants might encounter, ensuring a balanced approach that prioritizes participant welfare. Furthermore, the study refrains from advocating for any specific political cause.

The interview questions were designed to be non-leading and focus on experiences without introducing evaluative or judgmental elements. The guide included prompts about specific instances of content limitation or blocking, as well as reflections on emotional and practical responses to these experiences. This approach enabled participants to share detailed, personal accounts while keeping the conversation open and focused on their lived experiences. We started the interviews with warm-up questions about participants’ activism and motivations, before moving to their use of TikTok for political content (e.g., “Why have you chosen TikTok as a platform to display your political content?”), perceptions of the algorithm's influence (e.g., “How do you think TikTok's recommendation algorithm affects how many people see your posts?”), experiences with content moderation or suppression (e.g., “Have you ever suspected your political posts were limited or blocked by TikTok? How did you know?”), and strategies to increase reach or circumvent moderation (e.g., “Can you give examples of things you’ve tried to make more people see your videos or to prevent your posts from being blocked or limited?”). This open question approach allowed participants to shape the depth and direction of their responses, enabling them to share their accounts while respecting their autonomy in discussing sensitive issues. We opted to present statements as longer, verbatim quotes whenever possible to preserve authenticity. This approach ensures that the research is conducted with a commitment to neutrality and fairness, primarily focusing on capturing participants’ experiences without promoting or endorsing any political views.

Findings

TikTok and its potential for political activism

All participants in our study favored TikTok over other platforms for its viral potential through the recommendation algorithm and its connecting affordances. They described TikTok as the “de facto platform for activism, especially progressive activism,” emphasizing its ease in achieving wide reach and going “viral,” allowing them to spread awareness of urgent political issues to a large audience. They noted the algorithm's ability to personalize content feeds and how it enables activists to discover and “connect with other like-minded people” and engage with content that matches their interests, facilitating community building, networking, and mobilizing for their political cause. Interviewees also appreciated TikTok for its user-friendly interface and its inclusive nature, enabling individuals without specific skills, traditional qualifications, or platforms (like academic or public speaking opportunities) to participate in activism. They mentioned how TikTok's visually centered affordance enables authentic expression and can overcome spatial separation, allowing users to share personal stories and real-time updates, which supports grassroots activism by showcasing genuine voices and real-world impacts.

D: But if there's anything I think TikTok is ultimately good for, it’s that it puts faces to issues. It’s not like Twitter where someone could say ‘Help, I’m a Palestinian refugee’, but you’re just seeing text. Because no, you’re going to see a face, you’re going to see the environment they’re in, and that is a lot more humanizing to people. And I think that's why this sort of content has done so much better on TikTok than it has on other platforms.

Participants also appreciated the direct, personal interaction the platform facilitates through its connecting affordances like stitching, duetting, or going live together, explaining how this forces the audience to be exposed to social issues and different realities and encourages the mobilization and organization of actions, such as protests, boycotts, and fundraising.

TikTok and its challenges for visibility

One of the most significant challenges participants mentioned is the platform's content moderation practices, which most perceive as unpredictable, unreasonable, and biased, especially against their political content.

Experiences with content moderation

Content moderation often manifests through rapid and unexplained removals of interviewees’ videos, comments, stories, or account blocks due to alleged TikTok community guidelines violations. This strict moderation seems to particularly affect content related to Palestine, with activists noting that such content is often flagged or removed more quickly than other political content and without adequate justification. Multiple participants stated how this applied to “any kind of content related to Palestine” even if it was “not political at all.” One participant explained how they gave a historical lesson on Gaza but were still moderated because it contained footage of destruction. They also heavily criticized the moderation process for its opacity, unpredictability, and unfairness, as they regularly did not receive any clarification as to why their content breached TikTok's community guidelines.

G: Anything Palestinian, anything trying to raise money for the families, trying to evacuate, […] even just mutual aid [is being suppressed]. […] we have partners on the ground providing diapers, formula, food, sanitary kits, pacifiers, all that stuff, right? Videos that talk about that, are more willing to be suppressed.

F: It just actively says on the page [of TikTok's community guidelines] that what's allowed when it comes to talking about hate speech and discrimination and things is educational and documentary content raising awareness against hate speech, which I would argue my content always fell under. But that doesn’t seem to make any difference to it [the algorithmic moderation]. […] The way the algorithm functions [the algorithmic moderation], it doesn’t seem to relate to what's actually allowed and what's not.

L: TikTok has taken down a lot of my videos and it just says they go against community guidelines, but it won’t say what guidelines, which is very irritating. How do I know what guidelines that I'm breaking if you don’t tell me?

This uncertainty and lack of transparency also extend to the appeal process, which is seen as ineffective and frustrating, providing no chance for creators to receive a detailed explanation or to explain themselves. Creators also noted that successfully appealed posts, which were wrongfully moderated and therefore completely restored by TikTok, still are not getting any reach. They believed this was because after a video is restored, it will not get introduced into the algorithm anymore and has missed the time slot to go viral. Participants suspect that the appeals process is deliberately designed to be opaque, serving as an additional, subtle form of moderation that makes it more difficult for political content creators to get their message across. This sense of futility exemplifies what Divon et al. (2025) call platform gaslighting, where users are led to question their own actions rather than the system, eroding trust in moderation.

F: And our content fits into that category [of what is allowed according to community guidelines]; why we received a permanent ban isn’t clear. But we have no mechanism for appealing it, absolutely none. Like, you can send off as many of those requests as you want. Most of the time, at least in my experience and in the experience of a lot of people I know, […] the better idea is to just start another account.

C: Even if in context it makes sense, and you can appeal those things and try to get your video back. But the amount of time that they took it down may almost deaden it by the time it comes back. So, when the video goes back up, maybe an hour or two later, you missed your chance with the ‘For You’ page. So, people aren’t going to see the video anyway, unless they go to your specific page every day.

Overall algorithmic visibility moderation was perceived as a particularly challenging moderation practice, leaving participants uncertain about whether their posts have been affected and without an option to appeal. However, despite the lack of concrete proof, which prevents formal action, interviewees have noticed patterns and circumstantial evidence that suggest algorithmic suppression. They mentioned several indicators, such as a noticeable and abrupt reduction in views in the first few hours of posting compared to their other posts, their follower count, and overall engagement such as likes, shares, completion rates, and comments. Moreover, a disproportionate engagement-to-view ratio emerges, showing that some videos receive substantial engagement, such as likes, shares, completion rates, and comments relative to their views. Multiple participants concluded that this discrepancy implies that the content reached only a small audience and was held back from the FYP despite generating significant interest from the people it has been shown to initially.

J: I'll post multiple videos of different genres on astrology, and then activism vibes, and then actual Palestine or Congo related stuff. And I’ve noticed that Palestine related stuff: 400 views max. I have like 12,000 followers. […] If I'm posting something about activism or like U.S. political issues, then my video will maybe hit in the like 800 s to like low 1000 s after the first hour, but then it stops. It just stops abruptly. […] Astrology content is the only stuff that I get like thousands of views on, like consistently, even if it’s like 4000 views, that's definitely way more than 400 views. And so, you'll see that difference. And also, if you're in the creator program they'll show you your analytics for the video. And it’s like, oh, the video is doing good and then hard stop. Like just an immediately hard stop after like the first hour. And that's the main indicator of, okay, your video has been shadowbanned and stuff.

L: If I post other [non-Palestine] things, most of the time it’s more of an organic just straight like a line down as opposed to like a line and then a complete drop.

D: I have no verifiable proof of that. But what I do have is a lot of circumstantial evidence that seems to support that as a claim, which is to say, that most of the people who I follow, who a couple of months ago were doing very, very well in their views related to that subject [Palestine], are now struggling to break, you know, 1–2000 views, regardless of how many followers they have. And I am one of them.

K: You’ll see this spike […] where at first it'll get a lot of views and then it'll look like they'll finish their video processing and then it'll hit a wall and like tank again. […] The like to view, like to video or favorite ratio … views to likes is a really big one. So, if you get one in five, one in four, sometimes even one in three people who watched the video liked it, which is just an incredible ratio. And yet they will still not serve it out. That kind of a thing. So we'll look at those analytics.

Some participants even suspect their content or accounts being shadowbanned, a more serious form of algorithmic visibility moderation that makes their content, profiles or comments invisible to others while remaining visible to the creators.

L: I still have the screenshots of whenever I would send my friends my video, they'd say, ‘Oh, it’s unavailable’. And I'd go to my TikTok, and it'd be up for me. And this happened several times. It was just being hidden from people, even though the video was up.

These user experiences align with the previously introduced concept of platform gaslighting by Divon et al. (2025). Participants repeatedly described feelings of frustration, uncertainty, and mistrust. They try to understand the community guidelines but are left confused about what they did wrong, as platform statements remain vague and unhelpful.

Strategies to navigate algorithmic visibility on TikTok

To manage their algorithmic visibility, the interviewed political activists employ various content creation strategies, reflecting different visibility goals, primarily raising awareness and fostering solidarity, but at times also aiming for disruption. These strategies focus on circumventing and outsmarting the algorithmic systems used for content recommendation and moderation, with an emphasis on audience engagement as crucial for achieving algorithmic virality.

Circumventing and outsmarting algorithmic visibility moderation

To prevent or trick algorithmic moderation detection, participants modified their spoken and written language in captions, hashtags, subtitles, videos, and audio by using coded or misspelled words, emojis, and sign language, and by omitting specific terms in their metadata. They try to stay away from hashtags that specifically talk about free Palestine or free Gaza, as they have experienced that content with these hashtags is removed or suppressed. Instead, they use coded language that subtly hints at the topic but will not be noticed by the algorithm, allowing them to reach their audience without being limited in visibility.

H: Instead of censorship, I'll say S -E -N and like S -O -R, and then I'll put a ship [emoji]. It’s because the algorithm will pick up if I put the actual word, and the video will get taken down or capped at a certain number of views.

Interviewed activists also try to avoid algorithmic flagging by presenting their content in ways that lead the algorithm to interpret it differently. They conceal their true message by framing it in an entertaining or trendy way. They use popular sounds, viral hashtags, or trending topics to trick the algorithm into thinking their content is not political and, thus, reach a broader audience that might not engage in political content otherwise. These strategies echo tactics observed among other marginalized creator communities such as queer users, sex educators, and activists in authoritarian contexts who similarly engage in self-censorship, algospeak, and memetic reframing to navigate platform moderation (Are, 2024a; Bishop, 2019; Karizat et al., 2021), although in this case, such practices are shaped by the added pressures of geopolitical conflict and narrative erasure.

These vernacular adjustments are collective algorithmic imaginaries in action, shared, politicized repertoires for minimizing risk while maintaining meaning. While some of these adaptations may seem cosmetic, others carry deeper social consequences, particularly for political participation and representation. Algospeak, for instance, may appear to foster community identity through linguistic play, but it is primarily a survival strategy, a means of evading moderation and maintaining presence on the platform (Steen et al., 2023). Similarly, self-censorship emerged as a key theme in our interviews. Activists reported withholding ideas, language, or imagery out of fear that their content might be removed or suppressed. This phenomenon, often referred to as “chilling effect” (Penney, 2016), is notoriously difficult to measure because it leaves little trace. As Büchi et al. (2022) note, unspoken thoughts provide scant empirical evidence, making it challenging to assess whether these effects are widespread or disproportionately affect specific groups.

G: One of the fundraising videos I made that was to help a Palestinian family fundraise, I started off the video with, ‘Oh my gosh, Ariana Grande and Ethan Slater are getting married!’. And then in my video I go, ‘Okay, now that I have your attention …’, and then I write on the caption ‘oh, they're so cute together #arianagrande #wickedthemusical’ […] So then it gets to Ariana Grande fans ‘For You’ pages. […] Because I need to trick the TikTok algorithm into thinking that this isn’t a fundraising video, so that they don’t squash it.

D: If I don’t post content about Palestine, my views do seem to rebalance after about a week or so. I start to get more view counts that are in line with what I'd expect from my follower count.

Activists also alternate between posting political and non-political content, as they believe that sharing only political material causes the algorithm to suppress their entire account. This is typical for TikTok activism, which often features playful activism, characterized by memetic performances that blend entertainment with political messaging (Cervi and Divon, 2023; Cervi and Marín-Lladó, 2022). However, the use of American pop culture references, such as Ariana Grande and Ethan Slater, highlights their acute awareness of the cultural spheres they believe the algorithm favors.

Encouraging engagement and collaboration

Participants are very aware of the importance of engagement to enhance the algorithmic visibility of their content. As a result, they employ various strategies to boost interaction and gain algorithmic favor. They actively encourage viewers to like, share, and comment on their videos to generate discussions that simulate genuine conversations and ultimately enhance engagement metrics. Activists also prompt viewers to watch their videos fully, recognizing that complete views positively influence these metrics.

I: I’ve noticed when people do things like flower chains under these kind of posts, it doesn’t help as much as like, for example, somebody asking ‘what time is it for you?’ ‘How are you feeling?’ Then it opens up the box for people to have deeper and longer conversations, which then kind of makes the TikTok app be like, ‘Oh wait, people are really interacting with this video. Let me push it out to more people’. […] asking people to have conversations is the best way to get some kind of reach with these Palestine videos. And also asking people to watch it all the way through.

To further stimulate engagement, participants try to post content that catches people's attention, such as entertaining clickbait or rage bait intros, and simulate make-up tutorials by applying make-up to enhance their appearance. They also use dynamic camera movements and bubbly subtitles to draw attention to their political content. A lot of creators acknowledged that posting their original content in between pro-Palestine content helps boost engagement on their account, and therefore also for their pro-Palestinian videos, because their followers engage with them for their original content. Some creators also noted that authentic, relatable, and emotional content tends to attract higher engagement due to its humanizing quality. Additionally, creators often organize raffles, giveaways, and contests, acknowledging that these strategies significantly motivate audience participation. Some content creators have observed that their political posts get suppressed if they post frequently throughout the day, leading them to limit their posting frequency. Additionally, they have noticed that videos posted after a viral one tend to perform poorly, prompting them to share their most significant content early in the day.

H: I’ve had a number of videos go viral on TikTok, but then the videos after that, they always perform poorly. And I think it’s because a part of the algorithm boosts the content, but then it recognizes that the content is actually not what the platform supports. And so, then the account will be suppressed after that for any number of videos.

Many creators also maintain secondary accounts as a strategy to cope with their main accounts being flagged or suppressed by platform algorithms. Furthermore, creators often collaborate by seeking feedback from fellow creators before posting videos and promoting each other's content as part of a community effort. These efforts ranged from simple things such as commenting on each other's content to help with virality to more elaborate strategies such as the following example of “hacking the search bar” during the U.S. Super Bowl.

H: TikTok has a feature where every time you watch a post at the very top of the screen, there's a search bar. […] And if the same phrase is used in the comment section of the post over and over again, it will automatically fill that as a search. […] A bunch of TikTok users who are pro-Palestine basically brigaded this with the NFL's Super Bowl posts. They commented longer comments to make them seem more interesting and make them higher up in the theme. But they would all include the names of Bisan and Motaz [two Palestinian journalists]. And it totally worked with the search bar of the NFL Super Bowl post. The search was Bisan, the search was Motaz. It was amazing. It was very effective. So those are all kinds of things that we do. We know enough about the app. And then one person has a cool idea and then, you know, we'll implement it until they figure it out. And then we find some other thing we do.

This strong sense of community was present in all the interviews, with participants using collective language (“we,” “us”) to describe how they navigated content moderation and other challenges on TikTok together. This peer exchange reflects algorithmic gossip (Bishop, 2019): collective sense-making and knowledge-sharing under conditions of opacity. Overall, they demonstrated a nuanced understanding of algorithmic moderation, employing specialized terms like “algospeak” and linking their experiences to platform politics and political events, such as the U.S. Senate hearings on TikTok.

Discussion

Our study highlights how pro-Palestinian activists on TikTok navigate a highly specific sociopolitical and historical terrain that both enables and constrains visibility. The testimonies of these creators reveal how theories of algorithmic imaginaries, gossip, and gaslighting materialize in practice. Like other marginalized groups, our participants encounter opaque moderation, algorithmic volatility, and the constant need for strategic adaptation. Yet their experiences are uniquely shaped by a decades-long, politically charged conflict marked by foreign interests and a military occupation that extends into the digital sphere (Mahlouly and Erhaim, 2023; Rousselin, 2016; Roberts, 2019). As Elmimouni et al. (2024) and Mahlouly and Erhaim (2023) demonstrate in the case of Sheikh Jarrah, visibility challenges here are not only algorithmic but deeply politicized, with moderation frameworks that align with Western geopolitical priorities.

In our findings, this emerged as a clear cultural and political divide, with participants acutely aware that what was “actually allowed” or what the “platform supports” did not align with TikTok's stated community guidelines but instead depended on the political message. In response, pro-Palestinian activists on TikTok are in constant dialogue with the platform's recommendation system, experimenting with new formats and methods to maintain visibility and outsmart content moderation. Especially in times of war, these strategies often involve leveraging all available resources, including what Divon and Eriksson Krutrök (2024) term the platformed body, the strategic use of embodied performances, which in our interviews appeared as gestures, clothing, make-up, and emotional expressions, to enhance visibility.

With little verified information available, the interviewed activists rely on personal experience, circumstantial evidence, and collective sense-making to interpret how the algorithm functions. Concepts such as algorithmic imaginaries (Bucher, 2017), algorithmic folklore (Karizat et al., 2021), and algorithmic gossip (Bishop, 2019) describe this process. These remain crucial frameworks, and our case sheds light on how they intersect with recent work on platform gaslighting (Divon et al., 2025). Yet despite immense challenges and pressure, our findings reveal that the pro-Palestinian creators we interviewed continue to find creative ways around content moderation and find a considerable audience. The activists we spoke to experience the described symptoms of gaslighting, but they are not silenced, and their persistence suggests that algorithmic visibility moderation is not entirely successful. Asked about TikTok, their focus was first on the potential, and albeit through strategic labor and constant adaptation, they did find a big reach and voice on the platform and remained surprisingly positive about it. Paradoxically, their mix of wartime footage with playful content and formats, often performs well on TikTok, where they encountered a user base highly receptive to pro-Palestinian narratives. These findings echo Heřmanová et al. (2025)’s observation that once content passes moderation, “the algorithm loves the war.”

A recurring theme in our interviews was solidarity, as participants saw themselves as part of a strong, international community. Together, they exposed a structured regime of platform gaslighting, characterized by denial and inconsistency that normalize unequal visibility. They confronted it through mutual support and collective sense-making via algorithmic gossip, which, under these conditions, becomes vital coordination infrastructure rather than mere rumor. Frustrated by issues such as the appeals process and inconsistencies in, or absence of, platform responses, participants described successful coordinated efforts such as “hacking” the search bar as moments that gave them a sense of agency and efficacy. While they experienced content moderation as unfair, opaque, and biased, they adapted pragmatically: Instead of appealing moderation decisions, they often started new accounts or abandoned promising posts.

Furthermore, the activists we interviewed all had a relatively large following and turned out to be experienced content creators, familiar with advanced features such as insight metrics, which helped them collect evidence of algorithmic visibility moderation. This is where our study can add nuance to the concept of platform gaslighting (Divon et al., 2025). We conclude that much like gaslighting in domestic abuse, its effects are likely more severe when the victim is isolated or less digitally literate. Future research on platform gaslighting could explore how these effects play out in smaller or less connected communities, as such groups may struggle to find the reach and support networks we observed in our sample.

These findings point to another critical concern in platform governance: moderation bias. In line with previous studies, our interviews suggest that content moderation is experienced as biased, particularly against minority voices (Bacchi and Reuters, 2020; Brown, 2021; Cobbe, 2021; Steen et al., 2023). This is especially salient in the context of politically sensitive topics such as the Israel–Palestine conflict, where users report higher rates of suppression. Palestinians’ everyday experiences of war, destruction, and displacement can violate community guidelines simply by depicting reality. The activists we interviewed were acutely aware that Palestinian identity and narratives in general were being suppressed, such as the example of an educational video on Gaza's history that was removed for showing destruction. Similar concerns have emerged around trans and queer communities, where mere existence is politicized and often clashes with content moderation frameworks (York, 2022). This perceived unjust and opaque moderation puts pressure on groups already sitting at the intersection of multiple struggles, reducing the algorithmic visibility of marginalized content and hindering the spread of vital political information. It raises broader concerns about digital equity and the role of social media platforms in shaping public discourse. In the case of Palestine, these findings underscore the geopolitical dimension of moderation bias. Understanding it through concepts such as platform gaslighting and algorithmic gossip reveals how creators collectively interpret and resist systemic invisibility.

While the activists we interviewed remain visible and continue to fight back, their resistance comes at a cost. We must not underestimate the stress, exhaustion, and loss that accompany constant moderation struggles, nor forget the creators, and the narratives, already erased from visibility. As platforms consolidate power, align with political and authoritarian interests, and as technology becomes an ever more potent tool of oppression (Wynn-Williams, 2025), this dynamic grows increasingly alarming. Protecting these communities requires urgent regulatory oversight and stronger accountability mechanisms to ensure that digital platforms uphold transparency and political plurality. What happens to pro-Palestinian creators today concerns us all, for it can, and already does, happen to other marginalized communities in an increasingly authoritarian world.

Supplemental Material

sj-docx-1-pns-10.1177_29768624261427203 - Supplemental material for The quest for exposure: How pro-Palestine activists navigate algorithmic visibility and content moderation on TikTok

Supplemental material, sj-docx-1-pns-10.1177_29768624261427203 for The quest for exposure: How pro-Palestine activists navigate algorithmic visibility and content moderation on TikTok by Lukas Hess and Sarah Häusermann in Platforms & Society

Footnotes

Acknowledgements

We would like to thank all the participants who generously shared their time and experiences for this study. Their insights were invaluable to this research.

ORCID iDs

Lukas Hess

Sarah Häusermann

Ethical approval and informed consent

This study was conducted in accordance with the ethical guidelines of the Association of Internet Researchers (AoIR, 2019). All participants were fully informed about the aims and procedures of the study and provided their voluntary, informed consent prior to participation.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

Due to the sensitive nature of the data and to protect participant anonymity, the datasets generated and analyzed during this study are not publicly available. Further information may be available from the corresponding author on reasonable request, subject to ethical approval.

Other identifying information

There is no additional identifying information related to the authors, their institutions, funders, or approval committees that might compromise the anonymity of the review process.

Supplemental material

for this article is available online.

References

7amleh – The Arab Center for the Advancement of Social Media. (2025, January 29). Hashtag Palestine 2024: The war on Gaza, digital rights violations, and weaponization of AI. https://7amleh.org/post/hashtag-palestine-2024-en

Abbas

Fahmy

Ayad

, et al. (2022) TikTok intifada: Analyzing social media activism among youth. Online Media and Global Communication 1(2): 287–314.

Abdollahpouri

Mansoury

Burke

, et al. (2020) The connection between popularity bias, calibration, and fairness in recommendation. In: Fourteenth ACM Conference on Recommender Systems, pp.726–731. 10.1145/3383313.3418487

Abidin

(2020) Mapping internet celebrity on TikTok: Exploring attention economies and visibility labours. Cultural Science Journal 12(1): 77–103.

AoIR. (2019, Oktober 6). Internet research: Ethical guidelines 3.0. https://aoir.org/reports/ethics3.pdf

Are

(2023) An autoethnography of automated powerlessness: Lacking platform affordances in Instagram and TikTok account deletions. Media, Culture & Society 45(4): 822–840.

Are

(2024a) Dysfunctional’ appeals and failures of algorithmic justice in Instagram and TikTok content moderation. Information, Communication & Society 28(11): 1997–2014.

Are

(2024b) Flagging as a silencing tool: Exploring the relationship between de-platforming of sex and online abuse on Instagram and TikTok. New Media & Society 27(6): 3577–3595.

Are

Paasonen

(2021) Sex in the shadows of celebrity. Porn Studies 8(4): 411–419.

10.

Askanius

(2013) Online video activism and political mash-up genres. JOMEC: Journalism, Media and Cultural Studies 0(4). 10.18573/j.2013.10257

11.

Askanius

Uldam

(2011) Online social media for radical politics: Climate change activism on YouTube. International Journal of Electronic Governance 4(1/2): 69.

12.

Bacchi

Reuters

(2020, September 22). TikTok apologises for censoring LGBT+ content. Reuters. https://www.reuters.com/article/idUSL5N2GJ459/

13.

Banchik

(2021) Disappearing acts: Content moderation and emergent practices to preserve at-risk human rights–related content. New Media & Society 23(6): 1527–1544.

14.

Bandy

Diakopoulos

(2020). #TulsaFlop: A case study of algorithmically-influenced collective action on TikTok (arXiv:2012.07716). arXiv. 10.48550/arXiv.2012.07716

15.

Bennett

Segerberg

(2012) The logic of connective action. Information, Communication & Society 15(5): 739–768.

16.

Biddle

Ribeiro

Dias

(2020, March 16) Invisible censorship: TikTok told moderators to suppress posts by ‘ugly’ people and the poor to attract new users. The Intercept. https://theintercept.com/2020/03/16/tiktok-app-moderators-users-discrimination/

17.

Bishop

(2019) Managing visibility on YouTube through algorithmic gossip. New Media & Society 21(11–12): 2589–2606.

18.

Brown

(2021, June 7) TikTok influencer of color faced ‘frustrating’ obstacle trying to add the word ‘Black’ to his Creator Marketplace bio. Forbes. https://www.forbes.com/sites/abrambrown/2021/07/07/tiktok-black-creators-creator-marketplace-black-lives-matter/

19.

Bucher

(2017) The algorithmic imaginary: Exploring the ordinary affects of Facebook algorithms. Information, Communication & Society 20(1): 30–44.

20.

Büchi

Festic

Latzer

(2022) The chilling effects of digital dataveillance: A theoretical model and an empirical research agenda. Big Data Soc 9(1): 20539517211065368.

21.

Cervi

Divon

(2023) Playful activism: Memetic performances of Palestinian resistance in TikTok #challenges. Social Media + Society 9(1): 205630512311576.

22.

Cervi

Marín-Lladó

(2021) What are political parties doing on TikTok? The Spanish case. El Profesional de la Información 30(4): e300403.

23.

Cervi

Marín-Lladó

(2022) Freepalestine on TikTok: From performative activism to (meaningful) playful activism. Journal of International and Intercultural Communication 15(4): 414–434.

24.

Cervi

Tejedor

Blesa

(2023) Tiktok and political communication: The latest frontier of politainment? A case study. Media and Communication 11(2): 203–217.

25.

Charmaz

(2006) Constructing Grounded Theory. Vol. 1. London Thousand Oaks, California New Delhi Singapore: Sage Publications.

26.

Cobbe

(2021) Algorithmic censorship by social platforms: Power and resistance. Philos Technol 34(4): 739–766.

27.

Cotter

(2019) Playing the visibility game: How digital influencers and algorithms negotiate influence on Instagram. New Media & Society 21(4): 895–913.

28.

Cox

(2017) The source of a movement: Making the case for social media as an informational source using Black Lives Matter. Ethn Racial Stud 40(11): 1847–1854.

29.

Divon

(2022) Playful publics on TikTok: The memetic Israeli-Palestinian war of #CHALLENGE. Critical Meme Reader #2: Memetic Tacticality. https://www.academia.edu/91833110/Playful_Publics_On_TikTok_The_Memetic_Israeli_Palestinian_War_of_CHALLENGE

30.

Divon

Are

Briggs

(2025) Platform gaslighting: A user-centric insight into social media corporate communications of content moderation. Platforms & Society 2: 29768624241303109.

31.

Divon

Eriksson Krutrök

(2024) Playful trauma: TikTok creators and the use of the platformed body in times of war. Social Media + Society 10(3): 20563051241269281.

32.

Duffy

Meisner

(2023) Platform governance at the margins: Social media creators’ experiences with algorithmic (in)visibility. Media, Culture & Society 45(2): 285–304.

33.

Elmimouni

Skop

Abokhodair

, et al. (2024) Shielding or silencing?: An investigation into content moderation during the Sheikh Jarrah crisis. Proceedings of the ACM on Human-Computer Interaction 8(GROUP): 1–21.

34.

Eltantawy

Wiest

(2011) Social media in the Egyptian revolution: Reconsidering resource mobilization theory. International Journal of Communication 5: 1207–1224.

35.

Garrett

(2006) Protest in an information society: A review of literature on social movements and new ICTs. Information, Communication & Society 9(2): 202–224.

36.

Gillespie

(2018) Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. New Haven [Connecticut]: Yale University Press.

37.

Gillespie

(2022) Do not recommend? Reduction as a form of content moderation. Social Media + Society 8(3): 20563051221117552.

38.

Gonzalo

Medina

Rebolledo De La Calle

(2021) Politainment social audience and political engagement: Analysing Twitter conversations in Spain. Catalan Journal of Communication & Cultural Studies 13(1): 23–42.

39.

Gorwa

Binns

Katzenbach

(2020) Algorithmic content moderation: Technical and political challenges in the automation of platform governance. Big Data Soc 7(1): 205395171989794.

40.

Guinaudeau

Votta

Munger

(2020, Oktober 2) Fifteen seconds of fame: TikTok and the democratization of mobile video on social media. https://scholar.googleusercontent.com/scholar?q=cache:7Jmfu0F-u3QJ:scholar.google.com/+fifteen+seconds+of+fame+tiktok+and+the+democratization&hl=de&as_sdt=0,5

41.

Hautea

Parks

Takahashi

, et al. (2021) Showing they care (or don’t): Affective publics and ambivalent climate activism on TikTok. Social Media + Society 7(2): 205630512110123.

42.

Heřmanová

Eriksson Krutrök

Divon

(2025) The algorithm loves the war”: Ambivalent visibility in content creator practices during war. Continuum (N Y): 1–17. 10.1080/10304312.2025.2507777.

43.

Karizat

Delmonaco

Eslami

, et al. (2021). Algorithmic folk theories and identity: How TikTok users co-produce knowledge of identity and engage in algorithmic resistance. Proc ACM Hum.-Comput Interact 5(CSCW2): 305:1–305:44.

44.

Klug

Qin

Evans

, et al. (2021) Trick and please. A mixed-method study on user assumptions about the TikTok algorithm. In 13th ACM Web Science Conference, 84–92. 10.1145/3447535.3462512

45.

Köver

Reuter

(2019) TikTok curbed reach for people with disabilities. https://netzpolitik.org/2019/discrimination-tiktok-curbed-reach-for-people-with-disabilities/

46.

Le Compte

Klug

(2021) “It’s viral!”—A study of the behaviors, practices, and motivations of TikTok users and social activism. In Companion Publication of the 2021 Conference on Computer Supported Cooperative Work and Social Computing, 108–111. 10.1145/3462204.3481741

47.

EPH

Prasad

(2018) From Wall 1.0 to Wall 2.0: Graffiti, social media, and ideological acts of resistance and recognition among Palestinian refugees. American Behavioral Scientist 62(4): 493–511.

48.

Lievrouw

(2023) Alternative and Activist New Media. Vol. 2. Cambridge: Polity Press.

49.

Mahlouly

Erhaim

(2023) Pro-Palestinian activism: Resisting the digital occupation. In: Miladi

(ed.) Global Media Coverage of the Palestinian-Israeli Conflict: Reporting the Sheikh Jarrah Evictions. London: I.B. Tauris, 237–250. 10.5040/9780755649921.ch-014

50.

Matsakis

(2020, June 18) TikTok finally explains how the ‘For You’ algorithm works. Wired. https://www.wired.com/story/tiktok-finally-explains-for-you-algorithm-works/

51.

Monshipouri

Prompichai

(2017) Digital activism in perspective: Palestinian resistance via social Media. International Studies Journal (ISJ) 14: 37.

52.

Nicholas

(2022) Shedding light on shadowbanning [Preprint]. Open Science Framework. 10.31219/osf.io/xcz2t

53.

Nieland

(2008) Politainment. In: Donsbach

(eds) International Encyclopedia of Communication. Malden, MA: Wiley-Blackwell, 3659–3661.

54.

Office of the United Nations High Commissioner for Human Rights (2025, September 16) Israel has committed genocide in the Gaza Strip, UN Commission finds [Press release]. https://www.ohchr.org/en/press-releases/2025/09/israel-has-committed-genocide-gaza-strip-un-commission-finds

55.

Omnicore (2023, November 12) TikTok by the numbers (2023): Stats, demographics & fun facts. https://www.omnicoreagency.com/tiktok-statistics/

56.

Penney

(2016) Chilling effects: Online surveillance and Wikipedia use. Berkeley Technology Law Journal 31(1): 117–182.

57.

Roberts

(2019) Behind the Screen: Content Moderation in the Shadows of Social Media. New Haven: Yale University Press.

58.

Rousselin

(2016) Modern communication technologies and the extension of the territory of struggle: Conceptualising Tunisia’s jasmine revolution. New Media & Society 18(7): 1201–1218.

59.

Savolainen

(2022) The shadow banning controversy: Perceived governance and algorithmic folklore. Media, Culture & Society 44(6): 1091–1109.

60.

Scanlon

(2020, Mai 20) The app that knows you better than you know yourself: An analysis of the TikTok algorithm. Medium. https://chatbotslife.com/the-app-that-knows-you-better-than-you-know-yourself-an-analysis-of-the-tiktok-algorithm-be12eefaab5a

61.

Shankar

Dixit

Siddiqui

(2023). Are social media giants censoring pro-Palestine voices amid Israel’s war? Aljazeera. https://www.aljazeera.com/features/2023/10/24/shadowbanning-are-social-media-giants-censoring-pro-palestine-voices

62.

Shumate

Lipp

(2008) Connective collective action online: An examination of the hyperlink network structure of an NGO issue network. J Comput Mediat Commun 14(1): 178–201.

63.

Steen

Yurechko

Klug

(2023) You can (not) say what you want: Using algospeak to contest and evade algorithmic content moderation on TikTok. Social Media + Society 9(3): 20563051231194586.

64.

Tawil-Souri

(2012) Digital occupation: Gaza’s high-tech enclosure. J Palest Stud 41(2): 27–43.

65.

TikTok Help Center. (2023, November 12). How TikTok recommends content | TikTok Help Center. https://support.tiktok.com/en/using-tiktok/exploring-videos/how-tiktok-recommends-content

66.

TikTok Help Center. (2025, Juni 3). Community guidelines. https://www.tiktok.com/community-guidelines/en?lang=en

67.

TikTok Help Center. (2026, January 18). Content violations and bans. https://support.tiktok.com/en/safety-hc/account-and-user-safety/content-violations-and-bans

68.

Tufekci

(with Mediarep.Org & Mediarep.Org) (2017) Twitter and Tear Gas. The Power and Fragility of Networked Protest. New Haven: Yale University Press.

69.

Uldam

(2018) Social media visibility: Challenges to activism. Media, Culture & Society 40(1): 41–58.

70.

West

Blom

(2016) Explaining interviewer effects: A research synthesis. Journal of Survey Statistics and Methodology 5(2): 175–211.

71.

Wynn-Williams

(2025) Managing International Business: Open Access Edition I. London: AutoCognition Publishing.

72.

York

(2022) Silicon Values: The Future of Free Speech Under Surveillance Capitalism. London: Verso.

73.

Zayani

Khalil

(2024) The Digital Double Bind: Change and Stasis in the Middle East. New York, NY: Oxford University Press.

74.

Zeng

Kaye

DBV

(2022) From content moderation to visibility moderation: A case study of platform governance on TikTok. Policy & Internet 14(1): 79–95.

75.

Zulli

(2022) Extending the Internet meme: Conceptualizing technological mimesis and imitation publics on the TikTok platform. New Media & Society 24(8): 1872–1890.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

3.27 MB