Click farm as platform labor: Human bots in parasite platforms

Abstract

This article examines click farms in Brazil as a distinctive form of platform labor situated beyond dominant data work frameworks focused on artificial intelligence value chains. Drawing on multi-sited qualitative research conducted between 2020 and 2022, including platform observation, analysis of WhatsApp, Facebook, and YouTube content, and interviews with six women workers, the article analyzes the materialities, working conditions, and forms of organizing that sustain click farm labor. We argue that click farms operate as parasite platforms that depend on social media infrastructures while externalizing costs, risks, and material conditions of labor onto workers. Central to this arrangement is the production of artificial engagement, in which fake accounts and bots function not as deviations but as core features of click farm business models. The study situates click farms within longer histories of informal labor in Brazil, highlighting gendered survival strategies, parallel markets, and emergent forms of worker resistance.

Introduction

People who spend their days liking, commenting on and following profiles on social media platforms as a work activity at the behest of other platforms. These are click farm platforms, or simply click farms. The literature (Lindquist, 2018, 2021; Ong and Cabanes, 2019) points to their existence mainly in Southeast Asia, operating in spaces similar to call centers, simultaneously operating multiple mobile devices. This article contributes to global debates on data work (Muldoon et al., 2024) by shifting the analytical focus to data workers engaged in click farms and by redirecting the geographical lens from Southeast Asia to the Brazilian case. In Latin America, especially in Brazil, click farms present themselves as platforms, connecting clients who need “real followers” to workers, and are an important—and insufficiently explored—element of the platformization of labor (Weltevrede and Lindquist, 2024). While the major scholarship on data work focuses on labor aimed at training artificial intelligence systems, click farm platforms sustain an entire system dedicated to the production of artificiality. Rather than contributing to the development of technological systems, this labor is oriented toward generating artificial engagement on social media platforms.

This article examines click farms in Brazil as a form of platform labor, focusing on their materialities, working conditions, and associated parallel markets. We argue that click farms operate as parasite platforms in relation to social media infrastructures, within what has been described as a landscape of “platform trees” (van Dijck, 2021). The notion of the parasite, or parasitism, already appears in Marx's work (1976) as a phenomenon that exceeds the narrow frame of workplace exploitation, describing how capitalists benefit from the labor of others by appropriating surplus value at the expense of proletarian labor power. Building on this classical Marxist insight, this article conceptualizes parasite platforms by articulating it with contemporary platform studies, particularly the theoretical efforts of Weltevrede and Lindquist (2024). Drawing on Serres, they theorize the parasite as a third party positioned between social media platforms and users, creating “a new modality of exchange that remains dependent on the hospitality of the host” (Weltevrede and Lindquist, 2024: 4).

This conception is closely aligned with, yet analytically distinct from, Lata and Copolov's (2025) notion of “parasitic platform urbanism,” which was developed to describe how companies such as Uber depend on public infrastructures without reinvesting in them. In contrast, we conceptualize click farms as a slightly different form of parasite platform: one that draws on workers’ own infrastructures and exploits platform scams around social media platforms in order to generate value without producing any socially necessary technological or infrastructural development. Click farms extract value by appropriating and manipulating existing sociotechnical systems, while shifting costs and risks onto workers, who must themselves provide the material conditions of labor (including mobile phones, SIM cards, electricity, connectivity, time, and exposure to account bans or criminalization). Within the metaphor of “platform trees,” large social media companies function as hosts sustaining entire ecosystems of digital interaction, while click farms attach themselves to specific branches—such as likes, followers, and comments—to generate artificial activity. In doing so, they simultaneously undermine the credibility of engagement metrics and the legitimacy of digital markets, while reinforcing the centrality and profitability of the host platforms.

The parasitism of these platforms operates in tandem with the ongoing transformation of informal work in Brazil. Click farms in Brazil are best understood as a contemporary update of viração, a historically rooted mode of working-class survival structured around instability, and informality. Viração designates the provisional and adaptive character of work that sustains everyday life through “gigs” and other contingent activities, rather than stable employment relations, with a centrality of reselling products and services. The clickfarm workforce is predominantly female and shaped by long trajectories of informal labor, in which platform labor is combined with care and reproductive labor (Altenried, 2020; Tubaro et al., 2022). As Abílio (2011) shows, women's incorporation into the Brazilian labor market has long occurred under conditions of informality, low pay, and weak social recognition. Her analysis of direct cosmetic sales—where large companies outsource distribution to self-employed resellers without employment contracts—illustrates how precarious survival strategies are normalized. Click farms extend this logic into the digital economy, offering platform-mediated data work framed as “extra income” while relying on workers’ availability, flexibility, and self-provisioning of labor conditions. Rather than displacing informal work, digital platforms reorganize and intensify it. In the Brazilian context—where informal work continues to structure a large share of working-class livelihoods—click farms operate as parasite platforms that feed on historically consolidated practices of viração.

Building on Brazilian sociological traditions that understand informality as mobility across legal and illegal, recognized and unrecognized forms of labor (De Oliveira 2003; Rizek, 2006; Silva, 2011), we argue that click farms parasitize both social media infrastructures and workers’ own material resources. Much like earlier resale practices sustained by women's labor—from cosmetics to small-scale trading (Abílio, 2011; Pinheiro-Machado, 2017)—click farms extract value without generating socially necessary technological development.

They operate within broader circuits of para-platforms, in Steinberg's (2024) sense, as entities external to the platforms themselves yet integral to a wider circuit of activities organized around those platforms. This historical trajectory helps to situate contemporary forms of platform labor beyond the dominant frameworks of data work. Just as women relied on direct cosmetic sales as a mode of viração, digital workers today resort to similarly unstable and informal strategies to sustain their livelihoods. The emergence of parallel markets for the sale and purchase of fake accounts and bots, for example, can be understood as survival tactics within click farms, operating as small but significant fissures in platform power (Ferrari and Graham, 2021, Grohmann et al., 2022).

Methodologically, the research was conducted between October 2020 and January 2022 and adopts a qualitative, multi-sited approach grounded in messy digital methods (Dadas, 2016; Postill and Pink, 2012). Rather than privileging a single type of data or field site, we followed the infrastructures, practices, and discourses through which click farm labor is organized and experienced. This included systematic observation of four major click farm platforms operating in Brazil (GanharNasRedes, Dizu, Farmar Social, and SigaSocial), as well as sustained observation of Facebook and WhatsApp groups dedicated to click farm work. These online spaces were central for understanding both the operational logics of click farms and the broader para-platform circuits through which workers exchange information, tools, accounts, and services. In addition, we conducted semistructured interviews with six click farm workers, all anonymous, recruited through Facebook groups related to click farm labor. The interviews focused on workers’ life trajectories, working conditions, gender relations, and everyday interactions with digital technologies. Interviews were conducted online, recorded with consent, transcribed, and analyzed thematically. All research materials were securely stored in an encrypted folder. While interviews provided in-depth insight into workers’ experiences, they were complemented by extensive digital observation, which allowed us to trace how click farms function in practice, how tasks are distributed, and how risks, such as account bans, are managed by workers themselves. Finally, we analyzed 10 YouTube channels dedicated to click farm work in Brazil, which play a key role in recruiting workers, circulating tutorials, and framing click farm labor as “easy” or “extra” income. Taken together, observations of platforms, social media groups, video content, and interviews enabled us to examine click farms as part of a wider ecosystem of para-platforms structured around informal labor and resale practices. These methodological choices are oriented toward the article's central objective: to conceptualize click farms as a distinctive element of the platformization of labor, foregrounding materialities, working conditions, and emergent forms of organizing.

Accordingly, the article is structured as follows. It first situates click farms within broader processes of the platformization of labor, outlining their relevance for media and communication research. It then examines the materialities of click farm work, focusing on their dependence on social media infrastructures and the wider platform circuits in which they operate, including relations with clients. The subsequent section turns to working conditions, addressing gender relations, pathways into click farm labor, and its connections to longer histories of informal work in Brazil. The final empirical section analyzes the parallel markets for fake accounts and bots, interpreting them as both survival strategies and sites of fissure and organization within platform power. The conclusion synthesizes these findings, positioning click farms as a key site for understanding the intensification of platform labor and its implications for contemporary media practices. The main contribution of this article is theoretical, positioning click farms both within debates on platform labor beyond dominant data work frameworks and as an updated form of informal work in Latin America.

Click farms in the platformization of labor

We have argued elsewhere that platform labor functions as a laboratory of class struggles (Grohmann, 2025a). More broadly, platformization—understood as the growing dependency on digital platforms to organize work activities (Poell et al., 2021)—has generalized across sectors, involving heterogeneous platforms and worker profiles while relying on common mechanisms such as algorithmic management and data extraction as forms of capital (Grohmann, 2025b; van Doorn and Chen, 2021; Woodcock and Graham, 2019). Within this scholarship, data work has occupied a central place, alongside more visible forms of platform labor such as ride-hailing and delivery (Howcroft and Bergvall-Kåreborn, 2018; Vallas and Schor, 2020; Woodcock and Graham, 2019). Research across different regions has documented how data workers perform repetitive, low-paid tasks that sustain digital economies (Casilli, 2019; Gray and Suri, 2019; Jones, 2021; Roberts, 2019), including a growing body of work focused on Brazil (Braz, 2021; Kalil, 2020; Moreschi et al., 2020).

Much of the recent scholarship on data work, particularly in relation to artificial intelligence, has emphasized labor involved in training, moderating, and classifying data for AI systems (e.g. Muldoon et al., 2024). While this literature has been crucial for exposing the hidden labor behind AI value chains, it tends to overlook other forms of data work that do not directly contribute to the development of technological systems but remain central to platform economies. As Muldoon et al. (2024) note, there is data work beyond platforms. However, the case of click farms demonstrates that it remains analytically productive to understand such labor as platform labor, embedded in broader circuits of platformization. Click farms are deeply entangled with social media platforms and para-platform infrastructures, linking workers, clients, reseller markets, and automated tools in ways that reproduce core dynamics of platform labor despite operating at the margins of dominant AI value chains.

Building on Ekbia and Nardi's (2017) framework, we conceptualize click farms as a form of heteromation, or heteromated labor, defined as the “extraction of economic value from low-cost or free labor in computer-mediated networks,” which constitutes “a new logic of capital accumulation” (Ekbia and Nardi, 2017: 1). Rather than signifying automation in the sense of labor replacement, heteromation captures the expanding role of human work within processes commonly framed as automated, often producing the impression that there is no human in the loop. Instead of displacing workers, value extraction relies on the redistribution and reorganization of tasks across different labor arrangements, in which human effort remains central but increasingly obscured. In the case of click farms, heteromated labor takes the form of producing artificial engagement on social media platforms through the repetitive and intensive human work of a dispersed crowd of workers. The labor process involves clicking, commenting, and following accounts across multiple platforms, structured as fragmented and low-paid tasks. This dynamic highlights how heteromation is integral to understanding contemporary platform labor not as a process of technological substitution, but of continual worker recomposition. Click farms thus exemplify how platformization accelerates and intensifies labor fragmentation and taskification (Casilli, 2019), while simultaneously generating data and metrics that sustain the appearance of automation within digital economies.

Globally, an estimated 160 million workers engage in data work, the majority located in the Global South (Casilli, 2021). This distribution highlights the geopolitics of platform labor, whereby labor from the South sustains digital infrastructures, markets, and profits concentrated in the North. As Jones (2021: 13) argues, data work platforms are not a “phoenix from the South” but rather “another twist in our planetary labor crisis,” shaped by long-term processes of informalization, proletarianization, and declining labor demand. Situating click farms within this landscape allows us to expand prevailing understandings of data work and platform labor, showing how platformization also operates through parasitic arrangements that extract value from informal labor and social media infrastructures.

In Latin America, there is an intense market involving data work platforms, as research by Miceli and Posada (2021) shows, especially involving Venezuela, Argentina and Brazil. In 2019, Schmidt (2019) found that the majority of data trainers for autonomous cars in the world were from Venezuela. Ribeiro (2021) highlighted a Brazil–Pakistan–China production chain in which Brazilians were (under)paid to transcribe videos for TikTok for less than a dollar. According to Braz (2021), there are around 50 data work platforms operating in Brazil alone, divided into different subtypes, with headquarters in various countries around the world.

We classify click farm platforms as a specific type of data work platform, oriented toward producing artificial engagement for social media, rather than generating data for AI training or content moderation. We theorize the role of workers in click farms as “human bots.” They operate as if they were bots producing social media engagement, yet they are human workers performing tasks that mimic automation, in a characteristic process of heteromation. These workers sustain key segments of the social media economy, while being widely perceived as bots, illustrating how invisibility is central to platform business models that profit from heteromated labor (Raval, 2021).

The main click farm platforms in Brazil are GanharNasRedes, Dizu, FarmarSocial and SigaSocial, all based in Brazil, in cities such as Goiânia/GO and Santa Rosa/RS. Most of them define themselves as digital marketing companies focused on social media. Dizu, for example, presents itself as “an innovative startup that has found a way to fill the pain of people who want to become famous and the pain of people who are interested in earning extra income on the internet and thus remunerate themselves.” The statement equates the “pain” of those who want to become famous with that of those who need work platforms like these to manage their survival. This already reveals, in some way, a relationship between platform labor and the influencer/creator/celebrity industry.

The literature on click farms to date oscillates between analyses of their connections to the disinformation industry (Grohmann and Ong, 2024; Ong and Cabanes, 2019; Ong and Tapsell, 2022) and studies of illicit digital economies (Lindquist, 2018, 2021; Weltevrede and Lindquist, 2024). This scholarship emphasizes the creation and circulation of fake profiles as an integral and central component of the digital economy, framing click farms as factories of followers. This emphasis reflects the fact that workers often do not use their personal accounts when working for click farms; instead, they tend to create fake accounts that function as business accounts, used exclusively for click farm activities. Nemer and Sobral (2025) also address the role of heteromation in relation to artificial intelligence, highlighting its function as a human infrastructure for the circulation of disinformation through social media platforms. Lindquist (2021) goes so far as to describe click farms as “imposter infrastructures,” a characterization that highlights the centrality of infrastructural arrangements in these artificial economies. However, the perspective of work—and of workers—within click farms remains insufficiently explored in this body of literature.

Click farm platforms constitute a central node for scholarship in media, communication and platform studies, as they interconnect multiple actors within contemporary digital communication ecosystems. They link major social media platforms (such as Instagram and TikTok), click farm platforms that present themselves as digital marketing companies, a diverse range of clients—including politicians, influencers, and celebrities—who sustain and normalize these practices, and workers who perform these activities in exchange for minimal remuneration. In this sense, significant segments of digital communication in Brazil depend on a labor circuit in which click farms are deeply entangled with social media infrastructures. Taken together, this section shows how click farms crystallize key dynamics of platform labor, linking data work, informal economies, and parasitic value extraction within contemporary social media infrastructures. Against this backdrop, the following section examines the materialities through which these platforms operate.

Entering the platforms: Materialities of click farms

Understanding the materialities of platforms as means of production and communication (Williams, 2011) requires, simultaneously, examining how technological artifacts shape practices and processes, and situating these artifacts within the broader infrastructures to which they are connected. Approaches such as the notion of “platform trees” (van Dijck, 2021) or infrastructural mapping (Crawford, 2021) make it possible to locate platforms within wider sociotechnical systems, thereby positioning them in relation to ongoing processes of platformization.

Building on van Dijck's (2021) analogies in relation to trees, ¹ click farms are a kind of “parasite platforms” in infrastructural terms. This is because they have a logic of dependency on social media platforms, like a kind of platform scam (Grohmann et al., 2022). Click farms are only able to sustain themselves as platforms, with distinct business models (Poell et al., 2021), because they leverage the application programming interfaces (APIs) and platform-facing functionalities of social media services such as TikTok, Instagram, Facebook, and Kwai. APIs operate as layers of technical mediation that enable data circulation across systems and coordinate interactions between external para-platforms, automated tools, and user-facing features, without granting access to the internal infrastructures of social media platforms. Through these mediations, click farms connect their para-platforms to user profiles and engagement metrics, allowing clients to purchase followers, comments, and likes. This infrastructural dependency renders click farms structurally subordinate to social media platforms and their governance mechanisms, including rules for account creation, monitoring, and blocking—mechanisms that are neither neutral nor natural, but socially and politically constructed (Arriagada and Siles, 2020). Operating within these constraints, click farms organize labor through processes of heteromation (Ekbia and Nardi, 2017), in which human workers sustain the appearance of automated engagement. ² The result is the production of “artificial engagement,” generated through the coordinated interaction of workers, bots, and API-mediated systems, while remaining fully dependent on the infrastructures and metrics of host platforms.

As Weltevrede and Lindquist (2024) demonstrate, APIs play a crucial role in connecting click farm websites, bot infrastructures, and payment systems, enabling click farm ecosystems to scale and coordinate labor without direct integration into social media platforms’ internal systems. In this sense, APIs do not “infiltrate” social media platforms, but rather mediate interactions between para-platforms and platform-facing tools, facilitating the circulation of engagement as a commodity. This distinction clarifies how click farms operate as para-platforms: they remain dependent on the infrastructures and metrics of social media platforms while organizing labor, automation, and monetization through external systems.

The interfaces of the click farms are set up like websites and are aimed at working people. Before signing up, it is possible to simulate earnings with the platform. One of the buttons asks how many accounts you want to work with—an indication that fake accounts are possible. After signing up, you can add as many Instagram, TikTok or Kwai accounts as you like. However, there are minimum requirements—in order to comply with the social media platforms’ mechanisms for possible blocking. Some of the requirements are: the profile must have a human photo, a minimum of 60 followers, follow at least 10 people, have at least four publications, have a bio and a public/unlocked profile. One of the instructions is more specific in terms of minimizing the chances of being blocked by social media: “have a Brazilian name, username, and all photos must be of real people, no photos of objects, animals, games, sports, nature, adult content, random photos, etc.” In other words, with these instructions, click farms implies that workers end up creating fake profiles to work on the platforms, and try to create mechanisms in order to simulate authenticity in relation to the social media rules for blocking profiles. This constitutes another layer of heteromation, since the illusion of “real” accounts is sustained by the manual and repetitive labor of workers.

However, the platforms’ discourse is that they do not tolerate fake accounts. By “fake accounts,” platforms refer to accounts that are not linked to the workers’ own identities. One of the platform's rules is “only use real accounts” and states: “our system has advanced techniques for detecting fake accounts. If any fake account is detected, it will be deleted from the system and the points for the tasks performed by it will be refunded.” Thus, the platform publicly portrays itself as an organization that fights disinformation and fakes, but accepts and encourages what it calls “low-quality accounts.” In other words, click farms actively teach workers tips and tactics for creating “fissures” within social media platforms (Ferrari and Graham, 2021), a further characteristic of what makes them parasite platforms. These fissures involve encouraging workers to “scam” social media platforms (Grohmann et al., 2022) by exploiting their affordances, metrics, and governance mechanisms in order to generate artificial engagement. This dynamic also situates click farms within the broader disinformation industries (Grohmann and Ong, 2024). Platforms explicitly allow—and often instruct—workers to create and operate multiple accounts that are not tied to their own identities in order to boost artificial engagement on social media, whether by inflating profiles or commenting on posts as so-called human bots. Artificial engagement can thus be understood as “fake” not as a bug, but as a core feature of click farm platforms. Here, fake is used in a broad sense, encompassing both workers’ use of nonpersonal accounts, created manually, and the wider circuits of production, sale, and resale of social media accounts. In this way, click farms normalize and routinize disinformation-for-hire as a form of everyday digital labor (Grohmann and Ong, 2024). Thus, we argue that click farms profit from fake social media engagement regardless of how workers construct their accounts or whether they rely on third-party technologies.

These “tips” from the platforms give them an attribute of “mentoring” or what Soriano and Panaligan (2019) call “skill makers.” In the interface for the worker, there is a menu on one of the platforms called “tips and tutorials,” the main heading of which states: “tips for having fewer blocks on Instagram”—the main platform used. There are even YouTube videos for “creating high-quality accounts” and “avoiding/resolving blocks.” One of the tips is: “We recommend that you create accounts in a browser in incognito mode, preferably on a cell phone using the 3G/4G network. If you want to create more than one account, restart your phone before proceeding to create the next account. An amount considered safe based on feedback from our users is 30 actions per hour.” This means that by teaching tactics and fissures—as if they were circumventing the “rules of the game”—to workers, click farm platforms outsource the tasks of penetrating social media platforms to workers (Grohmann et al., 2022). Also, they build ways for their platforms’ business models to sustain themselves, because with fewer account blocks, there is more guarantee of satisfied customers.

Upon entering the platform, after registering their accounts, workers receive tasks through algorithmic management—similar to what happens on other digital labor platforms (Rani and Furrer, 2021). These tasks are generally to follow, like or comment on client profiles. When the worker clicks on the task, there are instructions on what to do. The platform warns them not to “undo the action”—for example, you should not “unfollow” the account after the action. In a statement of rules, one of the platforms asks: “Don't undo actions. We are honest with our users and allow them to earn money by performing actions and in return we ask that you also be honest with us and do not undo the actions that have been performed.” This is because, in the outsourcing of click tasks (Casilli, 2019), there is a platform-client agreement regarding the paid package of number of followers/likes/comments. Thus, workers are only paid for the tasks they perform and which can be checked by the platform. In other words, when a profile is blocked, the click farm cannot check it and the worker does not get paid for the task.

The value for each task is lower than other data work platforms, such as Amazon Mechanical Turk, Appen, or Lionbridge (Grohmann and Araujo, 2021; Moreschi et al., 2020) and has varied over time—an instability that also happens with other types of digital labor platforms. Until March 2021, one of the platforms paid BRL0.01 for each share. Then there was a change in the payment system and the amount became BRL0.006 per task—almost half of the earnings. The platform also states in the space where it communicates the amounts: “the amount may not seem much, but for the number of actions carried out per day you can have a good extra monthly income, considering that you only spend a few seconds carrying out each action!.”

In other words, the company's discourse reinforces the idea present in other platforms in the sector (Casilli, 2019) that the labor and the workers are unskilled and that it is just an “extra income”—something that also appears in the discourses of delivery platforms, for example, as shown by Anwar and Graham (2021). There is also instability regarding which social media platforms are included in click farms. In the beginning, workers could even be paid to dislike YouTube videos. Currently, Youtube is no longer available. The platforms available at the moment are Instagram, TikTok, Kwai and Facebook, with Instagram being the most used.

The production and consumption circuit (Qiu et al., 2014) of click farm platforms therefore involves, as we are describing, their materialities and infrastructures. This means considering that the existence of these platforms is related to the need for these services as commodities, in the Marxian sense. Based on our observations on the platforms, we looked at who the customers were, in a qualitative, random and nonstatistically representative sample, from which we can say that almost all of them are Brazilian. In this exploratory journey, we detected the presence of politicians, musicians, celebrities and soccer players. However, we consider their number to be small, almost anecdotal. Even though the sample has its limits, we can state that—from our observation—most of the clients are influencers/ content creators. This includes both verified and unverified profiles, with different reach and size. ³ However, we noted the strong presence of influencer profiles in the health sector, such as personal trainers and crossfit instructors. This means that there are links between the creator economy and click farms—as part of this platform labor circuit. Understanding click farms also involves knowing who the people who work for these platforms are.

Click farm workers

The literature on platform labor (e.g. Woodcock and Graham, 2019) and especially on data work (Casilli, 2019; Posada, 2022) has pointed out the difficulty of narrowing down to just one profile of those who work for these platforms, even with internal differences. For example, in Brazil, those who work for Appen and Lionbridge tend to have higher academic qualifications than Amazon Mechanical Turk (Grohmann and Araujo, 2021). In the case of click farms, the profile of workers is formed by people with a previous relationship to informal work, outside of the platforms, including as a garbage collector. This is also because, unlike other data work platforms, all the tasks on click farms are conducted in Portuguese, and with activities that can be carried out using cell phones. All of the click farm labor performed on the platforms we analyzed comes from Brazil, and we know this as Brazilian Portuguese speakers, because the Portuguese spoken in Brazil differs substantially in spelling and grammar from the Portuguese used in Portugal and in African countries.

There is also a strong presence of women working for click farm platforms. They are the ones who comment most on WhatsApp and Facebook groups and the ones who have the most “community management” roles on these social media. According to the Brazilian Institute of Geography and Statistics, women, young people and blacks have been the most affected in the labor market during the COVID-19 pandemic (Costa et al., 2021). In the interviews and observations, we met women who lost their jobs during the pandemic period and others whose husbands lost jobs and families needed more monthly income. “I started a little before the pandemic began. My son was very young and depended heavily on me, he was breastfeeding all the time, so I used that time to look for ways to earn income online” (Interviewee 3). In line with other research that points out the role of families in working for this type of platform in Latin America (Posada, 2021), updating the features of viracao in the region. By linking this phenomenon to gendered logics and the care economy, this research contributes to understanding how platform labor reconfigures structural inequalities and how the promises of “extra income” conceal long working hours and conditions incompatible with decent work.

Platform labor conducted remotely can renew and intensify forms of gender exploitation, in conjunction with domestic work and reproductive work (Altenried, 2020; Posada, 2022), based on the multiplication of work on different fronts. In the interviews, we discovered that all the workers are looking to increase their income, but not all of them manage to not work only through digital platforms but also have other “gigs.” Informality is exploited by platforms by capturing these workers in search of extra income through promises of high earnings. “When I started working online, I wasn’t doing anything else besides taking care of my son. I did have a secretarial job where I would occasionally fill in for someone, and there I continued using the platform on the office computers—it's been like that ever since” (Interviewee 1). These promises articulate work mechanisms that combine automated tasks with activities that require time and manual labor. All of this results in working conditions that are, at the very least, incompatible with the principles of decent work (Fairwork, 2022), because in an attempt to achieve the promised earnings, some workers end up practicing unhealthy routines and working unhealthy hours, with around 12 h a day spent on all the activities involved in the click farms. “I started in order to make money, and I wish these jobs were more consistent […] Now I think I can say that I’m a freelancer, but doing tasks that don’t require formal education” (Interviewee 2).

In addition to the actual tasks for the platforms, the workers highlight a broader circuit of work that involves creating profiles and working with bots (detailed in the next section), which means more working time. One of the workers interviewed said: “If you don't work hard, you won't win. So this image, ‘oh, I'm going to work little’, is misleading, because I work more than if I were working under a formal employment” (Interviewee 4).

Most of the workers interviewed said that they learned about click farms (which they call “websites”) through YouTube channels promising “extra income” and “easy income,” owned by people who present themselves as “digital marketing strategists” and “digital entrepreneurs.” The appeal of the video content is mainly focused on earnings, or promises of income and wealth. The strategies for attracting workers are, in addition to teaching how the tasks work—as skill makers (Soriano and Panaligan, 2019), to create hope in those who need to earn money. Some of the statements in these videos are: “It depends on whether you want to rest or earn money in your spare time”; “do something simple” or “turn into profit” what “you already do on a daily basis.” In this way, it reinforces the issue of informality, in a way disqualifying the task and, at the same time, putting pressure on those viewing the video to join click farms.

In Facebook and WhatsApp groups, we have seen how ads for “just click” jobs, together with the promise of “making easy money,” attract many workers with low levels of education who often don't even have their own computer or cell phone. They start working with a borrowed cell phone and, with the money they earn on the platforms, they manage to buy their first device. This is not an isolated example: reports of new acquisitions in the group—be it a cell phone, cosmetics or food—are frequent. In one case, a worker on these platforms was upset because other workers said that she just complains that she needs money soon and is anxious to get paid. When she was finally paid, she posted a photo of her fridge full of food, thanking the platform for the money, and “God” for being able to buy food, especially for having these platforms to work on. These reports of “thanking God” and the platforms are also very common in comments on YouTube videos. According to this worker, even with the problems generated by this type of work, these platforms help a lot when there is no other work.

Taken together, these working conditions make clear that fake is not a peripheral deviation but a constitutive element of click farm business models. Artificial engagement depends on workers’ systematic production and management of nonpersonal accounts, temporary emails, and backup profiles, transforming deception into an ordinary and everyday labor requirement. In this sense, click farm platforms organize and monetize fake practices, embedding fakery at the core of value extraction. Workers are thus positioned as “human bots,” performing these repetitive, standardized actions that mimic automated engagement while remaining fully responsible for the labor, risks, and ethical dilemmas involved. This arrangement exemplifies a parasitic mode of platformization, in which click farms extract value by attaching themselves to social media infrastructures and metrics, while externalizing costs to workers who must continuously multiply their labor, through account creation, device management, and risk mitigation, in order to secure minimal income. Parasitism here operates simultaneously at the infrastructural level, through dependence on social media platforms, and at the labor level, through the exploitation of workers’ survival strategies within heteromated technological systems.

The working conditions in click farms involve not only the direct relationship with the platforms, but the whole multiplication of work around them. This means all the work required behind the completion of tasks, such as creating accounts. As the pay per task is very low, workers resort to creating multiple profiles to manage their survival. In their view, this means ethical dilemmas when creating temporary emails and fake accounts to work for the platforms. One of the workers interviewed said that she has tactics for creating these profiles: she either creates accounts for celebrities or close relatives (and asks their permission), because she is afraid of being sued for creating fake profiles. Another worker says he would not have the courage to comment on and follow accounts with his own profile, and needs to separate the personal life on his Instagram from the “extra income” he earns from the platforms. Some even create fake faces created by artificial intelligence on sites like This Person Does Not Exist. These dilemmas are unresolved, as people need these activities for survival.

There are workers who work with up to 200 profiles at the same time, with several temporary emails. One worker said: “We do create fake profiles, otherwise you don't make any money” (Interviewee 5). It is often necessary to use more than one device—cell phone or desktop computer—to manage all the accounts. On one YouTube channel, a worker asks in the comments if these practices are legal, to which the YouTuber replies: “it's not illegal, but Instagram doesn't like it.” Thus, the biggest problem faced by workers is the blocking of their accounts by social media platforms. As a result, they do not get paid for their tasks. This is part of the multiplication of work in click farms (Altenried, 2020), as they will invariably need to work with an extra number of accounts in anticipation of blockages. One of the tactics that workers use is to keep backup accounts so that they can work while others are blocked. One worker says she has already had more than 200 reais deducted because of blocked profiles.

Another tactic used by workers, in addition to multiple profiles, is the automation of tasks using bots. This represents another dimension of fakery in click farms, in which workers, rather than manually creating fake accounts, outsource automation through bots, understood as fissures in the process. In other words, robots that can handle up to 300 accounts at the same time automatically perform the tasks of liking, commenting and following requested by platforms. In short: (a) clients buy click farms in search of “real followers”; (b) the tasks are outsourced to an underpaid crowd of workers; (c) tired of the conditions, the workers outsource the service once again, this time to bots (precisely what the clients wouldn't want to buy). Then, with the use of bots, if there are blocked accounts, at least they haven't been handled manually with people's labor. One of the workers said: “There was a time when it was wrong to use bots. We were blocked. Nowadays, we can only make money using bots. Those who only do manual labor aren't making it.” Thus, the use of fake accounts and bots is part of the day-to-day work on click farms. But they are the gateway to parallel markets, especially in WhatsApp and Facebook groups, which end up engendering forms of worker organization.

Parallel markets and forms of organizing

Fake accounts and bots are not born in a vacuum. They are often the product of parallel markets in WhatsApp groups, with titles like “TikTok and Instagram Sales,” “Growing on the platforms” and “Official Insta Earning Money.” They are part of the broader circuit of platform labor, operating as para-platforms that support and reinforce click farm platforms. WhatsApp groups are the infrastructures that facilitate these parallel markets, as places for trade, communication and organization between workers. They are a synthesis of what Soriano and Cabañes (2020) call “entrepreneurial solidarities,” and have taken place in groups of platform workers from various sectors (Grohmann et al., 2023). There is, at the same time, a feeling of mutual help between the members of the groups and reinforcements of an entrepreneurial neoliberal rationality, as a contradiction that cannot be resolved (Harvey, 2017). Workers share both tips for working and advertisements for accounts and bots. Thus, both solidarity networks are formed as emerging forms of organization, as well as possibilities for turning groups into yet another facet of the market.

One of the ads, for example, says: “Good evening, friends! Buy good Instagram accounts for R$1.50. I have clients who have been running for two weeks without being blocked. I have referrals! Accounts come with bio, 100 followers, 5 photos, like on all.” The sellers promise high performance, and sell everything from “empty accounts” (for around BRL0.70) to what they call “high standard accounts.” The more followers, photos and details the account has, the “higher standard” it is, and consequently the harder it is to block by social media platforms. These accounts are used solely for the purpose of following and commenting on profiles for click farms. In other words, the parallel market serves the work circuit on these platforms.

In the interactions, there are interactional dynamics that relate to the historical informality of labor in the country, such as viracao, of a core practice of reselling products and services. One example is this interaction for the sale of accounts: “zeroed accounts for R$1, interested parties call in private,” “selling for 90 cents,” “I'll do 80 cents then,” “I'm doing it for R$0.65.” There are also ads for the “resale” of accounts and for workers to become potential resellers, in an update—through platformization—of the historical resale activity in Brazil (Abilio, 2014).

Many of the fake accounts on the parallel market are women, as this ad shows: “Real, female names, without all those numbers and symbols.” There are also photo packs for building these profiles, usually of women: “pack with more than 90 photos, female profile, not a celebrity or blogger, always generates a lot of engagement, 25 cents.” In other words, there are various gender dimensions involved in working on click farms—from the women workers to the circulation of fake profiles through illegal markets (Pinheiro-Machado, 2017). This also shows how this work circuit involves several layers beyond the worker–platform relationship itself, as para–platform relationships.

The sale and purchase of bots follows the same logic as the fake account market. But in addition to the bots, mentoring and tutorials on how to use the bots better are sold by skill makers (Soriano and Panaligan, 2019) circulating in WhatsApp groups. In the groups, workers discuss which are the best bots and the best performers. They also share screenshots and videos of the bots working, to show other workers that “their bot” is outperforming the others—in a typical example of entrepreneurial solidarity, that is, a mix of competition between workers (“mine is better”) and building networks for help. The sharing of screenshots—especially remuneration—also happens in platform work in other sectors, such as transportation and delivery (Author, 2021).

The use of bots has recently been reappropriated by the click farm platforms themselves, which have created their own bots and produced tutorials and advertisements so that workers don't resort to parallel markets, but buy from the platform itself because it is more “reliable” and “official” than WhatsApp groups. This exemplifies how there is an intense game between the possibilities of agency and fissures in algorithmic power (Ferrari and Graham, 2021) on the part of workers and the logics of incorporation/reappropriation on the part of the platform power.

However, this does not mean that click farm workers are completely disjointed in organizing terms. On the contrary, in their own way, they show that “struggles against platform capitalism” (Woodcock, 2021) are taking place in the different sectors of the platform economy. When the click farms lowered their minimum rates in March 2021, the workers instituted the first click farm strike in Brazil (called by them the “strikes on the websites”). The protests began in WhatsApp groups and were supported by YouTube content creators—which also shows their contradictory role in relation to the reproduction of entrepreneurial rationality. This creator shared link for workers have discussion about the strike and explained the reasons for the stoppage: “if we don't act today, tomorrow the minimum rate will be BRL$0.0001.” In the comments on the channels, workers share organizing tactics, such as “let's stop following everyone. When the buyers see that they're losing followers and complain on the sites, they'll go back to their previous scores.” The workers also recognize the need for collective organizing: “I finally found a channel that makes us who work with these platforms more united. We have to stick together,” “top man, let's paralyze these parasites.” This demonstrates that even in unlikely scenarios of worker organizing and mobilization, workers create fissures and possibilities, however short-term, of agency in relation to platform power. It is also noteworthy that workers themselves recognize these platforms as “parasites,” in the sense of parasite platforms.

Thus, WhatsApp groups outline both forms of organizing and are the gateway to parallel markets, crossing borders of illegality and intensifying the informality of work in the broader circuit of platform labor. The use, sale and purchase of fake accounts and bots are part of workers’ strategies in the struggle to manage their survival, in a context of the deep web of platform labor in Brazil. The structure of parallel markets also shows that click farms, as parasite platforms, are updating informal work in Brazil, with a central role for the resale of products and services, much as in earlier periods.

Conclusion

This article has examined click farms in Brazil as a distinctive form of platform labor beyond dominant forms of data work within AI value chains, showing how they intensify, update, and reorganize historically entrenched forms of informal work through processes of platformization. By analyzing click farms across three interconnected dimensions, materialities, working conditions, and forms of organizing, we have demonstrated how these platforms operate as parasite platforms that depend on social media infrastructures while externalizing costs and risks onto workers. Click farms rely on workers’ own devices, connectivity, time, and emotional labor to sustain the production of artificial engagement, positioning fake not as a bug, but as a core feature of their business models. In doing so, they reveal how platform economies continue to thrive through the reorganization of informality in Majority World countries.

Theoretically, the article contributes to platform labor scholarship by extending debates on data work beyond dominant frameworks focused on artificial intelligence training, moderation, and classification. Conceptualizing click farms as heteromated labor allows us to foreground how human work remains central to processes framed as automated, even when no technological system is being trained or improved. Workers function as “human bots,” with similar tasks as bots, performing repetitive and standardized actions that simulate automation while remaining fully responsible for managing multiple accounts, avoiding detection, and absorbing the consequences of platform governance such as account bans and unpaid labor. The concept of parasite platforms helps clarify this arrangement by capturing how click farms attach themselves to existing platform infrastructures and metrics, likes, followers, comments, without contributing to their maintenance, while simultaneously reinforcing the economic and symbolic centrality of major social media platforms. This parasitic logic operates at both infrastructural and labor levels.

The concept of parasite platforms contributes theoretically by positioning click farms as entities that survive on the coerced “hospitality” of large social media infrastructures, while simultaneously exploiting workers’ physical infrastructures and labor power to generate value without producing socially necessary technological development. Within this configuration, the figure of the “human bot” emerges as a key concept for understanding these new labor processes in platforms. Click farm workers embody the central contradiction of heteromation: manual, repetitive, and exhausting human effort that is deliberately obscured in order to simulate automation. Studying parasite platforms therefore requires looking beyond “Big Tech” or dominant AI value chains firms to examine the wider circuits of para-platforms and parallel markets that sustain the digital ecosystem. At the same time, forms of worker resistance, such as “website strikes” and the tactical use of bots against the platforms themselves, demonstrate that even under conditions of extreme invisibility and fragmentation, fissures in platform power persist.

Beyond the Brazilian case, these findings speak to broader international debates on platform labor by highlighting how platforms, and their wider circuits of para-platforms, reorganize informal work into arrangements that are often more precarious than those addressed in the dominant platform labor literature. While click farms have often been associated with Southeast Asia, this article shows how similar dynamics happen in Latin America through historically specific configurations of informal work, gendered labor, and resale practices, now updated in the form of platforms. This Brazil-based research can be extended through comparative work on labor markets in other Majority World regions, particularly across Asia and Africa, where long-standing histories of informal work are likewise being reconfigured through platformization. Situating click farms within traditions such as viração highlights that platform labor reorganizes and intensifies informal economy, particularly in contexts where informality has long structured working-class survival. More broadly, the article invites scholars to rethink platform labor as an ecosystem that includes para-platforms, parallel markets, and possibilities of worker fissures alongside dominant platforms.