The presentation of The Bots is added to by the nine-part work Abuse Standards Violations, which marks the beginning of Eva and Franco Mattes’ research on the subject of content moderation. It takes as its theme the issue of the morality of social media and the tech giants associated with it.

Nine wall plexiglass frames, filled with insulation materials, present corporate guidelines, for example excerpts from the Facebook Community Standards, which are not intended for public viewing but for internal purposes only. The companies that have produced these guidelines are almost all unknown, as they wish to remain anonymous. Most of the time even the moderators themselves do not know who their employer is – one of them told Eva & Franco Mattes: ‘I’m pretty sure I work for Google’. The guidelines against violations of abuse standards set moral boundaries for what the companies consider questionable content on social media, laying down what is defined as racist, hateful, controversial, terroristic, pornographic or violent and thus to be removed. ‘Clean’ or ‘OK to show’ refers to images that are considered proper and therefore can circulate on social media, like ‘Shirtless but wearing pants or shirts (and not more than the top band of their underwear is visible)’; ‘inappropriate’ images may include politics and controversial social issues and so should be filtered. ‘Safe’ content includes fine art and celebrity gossip. Despite set guidelines, there is confusion as to when content should be removed, and who gets to decide what to remove. At this point interpretation made by humans is required, an algorithm-based assessment being insufficient.

The policies of large social media platforms change daily. They adapt to current social and political events. Since most IT companies are based in California, they mostly follow the guidelines of US laws and US ‘morality’, yet strive to be sensitive to local and culturally specific morals. The difficulty lies in exercising content moderation for all cultural contexts in a way that avoids the danger of allowing cultural biases to become political interpretation.

For The Bots, Eva & Franco Mattes collaborated with investigative journalist Adrian Chen and actors and actresses Irina Cocimarov, Jesse Hoffman, Jake Levy, Alexandra Marzella, Ruby McCollister, and Bobbi Salvör Menuez. They present anonymous testimonies from content moderators who have worked for Facebook in Berlin. Six videos have been created. In the room, visitors observe raised tabletops that form a minimalist installation. These tabletops are a reference to the furniture found in the Berlin moderation centre where the interviewees worked. The videos become visible to viewers only when they step behind the erected barrier and look behind the surface of the work.

What do we know about the mechanisms and regulations of social media channels that we use daily? Which contents remain visible and which are filtered out? And are there clear guidelines according to which content is deleted?

The films were executed with the typical aesthetic and features of online make-up tutorials. The statements in the films are derived from investigative research and interviews conducted with numerous witnesses employed as service providers for Facebook. The films were interpreted by actors so as to anonymise the statements of the content moderators. They perform the role of influencers addressing their followers directly. They recorded the videos using smartphones, for which reason the images are in portrait format. Advice on make-up products alternates with distressing descriptions of moderators’ work.

Content on social media channels is subject to restrictions and is thus scrutinised and monitored. Platforms claim to regulate their content through community guidelines. Some channels like Telegram also allow uncensored and problematic content. The guidelines cannot prevent thousands of ‘prohibited’ content from being posted online daily, however: violence, sexual assaults, hate speech, terrorism and pornography are just some of the categories of unwanted content on social media. Most of this content we cannot see, as it is deleted beforehand. This critical review is always carried out by human beings, i.e. it is not an automated cleansing process performed by algorithms. While programs filter content that appears to violate the guidelines of the respective platform, they cannot usually provide an independent interpretation of a post’s specific context.

In their work The Bots, Eva & Franco Mattes explicitly draw attention to the fact that critical content is seen and processed in large quantities by individuals. They are not bots, nor programs, but humans. They are called ‘content moderators’, and their profession falls within the category of ‘unregulated’ jobs that have emerged with the rise of tech companies (e.g. Amazon’s Mechanical Turk).

In the case of crowd-sourced job placement, content moderators often do not know themselves which companies they are working for. They are employed by so-called contractors who broker between tech giants like Google, Meta, YouTube, Twitter and the employees. In this way, the anonymity of the companies is preserved, their legal responsibility minimised and protected by non-disclosure agreements. Working conditions are neither publicly debated nor politically regulated. Services are governed by temporary employment contracts and are minimally paid.

Thanks to investigative journalism, reports on misconduct have nevertheless repeatedly reached the public domain in recent years. Journalist Adrian Chen was the first to shed light on the topic with his 2014 article in Wired titled ‘The Laborers Who Keep Dick Pics and Beheadings Out of Your Facebook Feed’. Eva & Franco Mattes have been collaborating with Chen for years.

One of the main problems is that content moderators have to review thousands of posts daily before deleting them. Beheadings, child pornography, explicit violence of all kinds, fanatical hate speeches, and many other expressions of the depths of human depravity, as well as the sheer flood of banal uploads, leave in their wake trauma and profound disturbance in people.

While guidelines exist on the classification of objectionable content, these are not made public and must be kept secret by content moderators. The regulations are subject to daily changes. In order to quantify the moderators’ performance, a minimum deletion rate of 95% of the contributions must be achieved, otherwise the employee is sacked. According to anonymous statements made by employees, workers are monitored and under intense pressure to perform.

In many cultures, the rules are adapted to fit the locally prevailing conception of morality. Also, content moderation is often carried out by workers in the Global South, Asia and former colonies. The reason for this, apart from unregulated labour law, is a command of Western languages and an awareness of Western moral sensibilities.

However, even moderators are not objective filters. Despite guidelines, the process is subjective, influenced by individual interpretations. Content is removed, for example, when it is deemed politically or ideologically inappropriate. One’s own political leanings can potentially influence moderation decisions.

Through their choice of aesthetic, Eva & Franco Mattes create a deliberately jolting break with the content. They employ the staging of make-up tutorials for their artistic work. Political content is camouflaged to avoid censorship. This approach derives from activists who use this method to bring political messages and human rights violations in autocratic states to the public’s attention. It was the young TikTok user and activist Feroza Azis who filmed herself putting on make-up a few years ago in order to circumvent the censorship of the Chinese government. This enabled her to speak freely about the systematic repression and surveillance of Uyghurs in northwest China before she was blocked from the platform.

At the same time, make-up as a subject is to be understood symbolically. As the artists themselves say: ‘Make-up is a way of concealing imperfections in our faces, not much different from content moderation, which beautifies the surface of the internet by removing unwanted content.’

Personal Photographs November 2007 is a self-contained network between two Raspberry Pi microcomputers connected by cables and constantly exchanging files with each other. The cables and cable trays create a temporary site-specific sculpture. As the title of the installation suggests, 101 personal photos of the fellow artists circulate in the closed system. The image files, however, remain deliberately invisible to the visitors – images without viewers, yet always there. Like most images nowadays.

The installation is based on the code developed with David Huerta and available on the Github open-source platform. Huerta is a digital security trainer at the Freedom of the Press Foundation, where he works on ways to train journalists to take advantage of privacy-enhancing technologies to strengthen a free press. Open sourcing allows developers and artists worldwide to use, extend and adapt the code.

The image files circulating in the Personal Photographs system were taken by Eva & Franco Mattes in November 2007. The selection gives an impression of the huge quantity of images that accumulate on mobile phones, computers and data centres as communication and interaction increasingly take place in the form of digital images and are uploaded in vast quantities.

Since the public internet first emerged, there have been significant phases of development. Initially used as a read-only instrument – i.e. purely for accessing information – it became an interactive communication network. Direct user participation became possible from 2004 on with the introduction of Web 2.0, enabling anyone to generate and publish content themselves. In subsequent waves, image platforms became the trend, replacing each other in ever more rapid succession: the image host Flickr (2004), the photo blog WordPress (2005), the social network Facebook (2004), the microblogging service Twitter (2006), Instagram (2010) or TikTok (2016) as well as instant messaging apps like WhatsApp (2009), Snapchat (2011) and BeReal (2020).

The vast majority of images nowadays do not exist in the form of printed photographs, hung on a wall or featured in a book, rather as ubiquitous files that are constantly copied and transferred between devices, from one data centre to another, via miles of cables or through thin air.

The voluntary participation of all users worldwide offers the few global internet corporations the possibility of using the totality of published content as data sets. On the one hand, little awareness exists of the corporations’ access to the data. On the other hand, the accumulation of information by the small number of major corporations is so frighteningly high that they can use big data management to make predictions about collective behaviour – and also to exercise control over societies. Politics and democratic structures lag behind.

Digital images reveal additional information via the metadata: the date and time the picture was taken, geographical coordinates, but also details about the technology used. Thus, when images are shared, additional information is unwittingly passed on to the public. Furthermore, since its creation, social media content has also been used as data sets for machine learning without the knowledge of the users.

While users upload content on social media for entertainment and leisure, this accessible information is used and monetised by companies to generate revenue with no concern for authorship.

Long before the advent of social media, Eva & Franco Mattes explored the sharing of personal information. In their performance titled Life Sharing, which took place from 2000 to 2003, the artists published all the contents of their computer: all their artworks, as well as private material – including emails, texts, photos, and bank statements – were freely available for viewing through their website. Considered a radical – and paradoxical – gesture at the time, today this act of excessive sharing is perceived as acceptable, even desirable, on social media.

Through their installation Personal Photographs, the artists revise this practice and exhibit a private archive to which outsiders have no access. Only the support structure, the hardware, remains in the space as a sculptural manifestation.

What is revealed here is not their private space, rather the infrastructure of data. The materiality creates a presence in the space, reminding us that digital content and images require a material infrastructure to be stored, sent and shared. The physical fragility of digital networks is transformed into sculpture. Adapting to the pre-existent architecture, it influences the way visitors move in the space. In this case it only channels physical movements, but of course technology shapes our behaviour, emotions, memories, expectations, fears and dreams, too.

We thank OBO Bettermann.

Faina Yunusova is concerned with the conventions of the internet and the emergence of communities that contribute to the collective construction of digital phenomena. She identifies and uses as many as possible of the factors that in social media generate success and high user numbers.

The digital performance #SugarMacht created for the Frankfurter Kunstverein begins on Instagram and finds its extension in the exhibition space. Yunusova cites a variety of currently successful aesthetics that can be seen as a backdrop for self-staging in social media. The purple light, the ring light for better self-portraits, the intimate atmosphere of a bedroom. The bed stages a supposedly private sphere of the artist that only serves to present herself to the public. The neon writing hanging from the ceiling with the hashtag #SugarMacht is an invitation from the artist to the visitors to participate in the performance and become part of her work.

The first screen shows a compilation of content produced and posted by the artist herself. Here, Yunusova stages herself as an influencer who performs TikTok trends, thus inviting the audience on site to participate and imitate her. Instagram and especially TikTok reinforce the practice of remix. Mashups and challenges are generated based on the repetition and interpretation of others’ specifications. Her mise-en-scène is a fake, practiced as a convention of online self-representation. The artist uses her body as a placeholder for collective imaginings, but through a variety of filters, music and comedic dislocations it is used as a staging of a digital person.

The space visible in the digital presentation and potentially misjudged as reality is actually entered by visitors to the exhibition and thus made usable again as a backdrop. Yunusova creates a loop of self-reflection between digital and analogue, in which the staged space is not only viewed via the mobile phone interface, but is offered as a real stage on which the visitors now also perform themselves. If you post a picture or a video on Instagram under the hashtag #SugarMacht, you become part of the installation – the audience can view themselves in real time on a screen that is connected to the internet and randomly displays posts that have been shared under the hashtag.

Through her own language based on dancing, music and self-parody, the artist shows the playful nature of social media (#Sugar) and reveals how the spread of trends builds collective identity, actively and increasingly merging an online community. Simple interaction mechanisms of social media are revealed. What appears banal suggests the dependence on self-dramatisation and the power of mass production in the making of content that can result from the use of social media (#Power – in German #Macht). Those who define themselves as influencers and believe they have power to “influence” people are themselves influenced by the mass culture of others. Thus, the hashtag #SugarMacht indexes the ambiguity of social media.

Faina Yunusova (*1991, Tashkent, UZ) currently lives and works in Offenbach am Main (DE). Her work explores digital presence on social media and the creation of virtual identities. She has been studying at the Hochschule für Gestaltung Offenbach (DE) since 2017. She previously studied at the Stroganov Moscow State Academy of Arts and Industry in Moscow (RU) before being expelled from the university for her internet activism. Among others, Faina Yunusova has exhibited at the following institutions: Basis Projektraum, Frankfurt am Main (DE), Rumplmayr Neukirchen, Sägewerk, Neukirchen (AT), Academy for interdisciplinary processes, Offenbach am Main (DE), Zwo Gallery, Vienna (AT).