AoIR2022-paper submission_ID_644_final.docx Selected Papers of #AoIR2022: The 23rd Annual Conference of the Association of Internet Researchers Dublin, Ireland / 2-5 Nov 2022 ALGORITHMIC JUSTICE FROM BELOW?! Aikaterini Mniestri University of Siegen Sebastian Randerath University of Bonn Abstract Datafication has become an integral part of users’ social media experience and the political economy of platforms. Given that the political economy of social media and advertising platforms is largely driven by participatory culture (Burgess & Green, 2018), platform companies are incentivised to open up the data vault to users, just a smidge, allowing them to access a series of metrics, neatly organized in analytics interfaces referred to as ‘creator studios.’ Creator studios are part of a larger assemblage of machine readable metrics and data analytics on the back end of a digital platform. While it has become common practice for creators to make use of this feature to grow their business, creators have also appropriated it to engage collectively in “algorithmic gossip” (Bishop, 2019) around social issues. In a number of cases, we observe how algorithmic gossip can become instrumental in the hands of minoritized groups (D’Ignazio & Klein, 2020, p. 26). Particularly, following the Black Lives Matter mass protests in 2020, creators noticed that their posts were less visible after they used their platforms to report on the movement and the protests. As more and more creators took screenshots of their analytics and posted them on Instagram, YouTube and TikTok, they were able to force a response out of the company. In response, Tik Tok called this a ‘glitch’ and launched an incubator to boost talent from the black community. Creatives had to apply for one of one hundred spots in the incubator meeting particular criteria. To be eligible, creators had to be over 18, based in the United States of America, and number a minimum of 10,000 followers. Having chosen a hundred creators for the “TikTok for Black Creatives” incubator program (TikTok Newsroom), the company made a commitment “to elevate and amplify their voices” by supporting their careers and developing their personal brand over a period of three months. Additionally, for this Mniestri, A.; Randerath, S. (2022, November). Algorithmic Justice from Below?!. Paper presented at AoIR 2022: The 23rd Annual Conference of the Association of Internet Researchers. Dublin, Ireland: AoIR. Retrieved from http://spir.aoir.org. project, TikTok has collaborated with MACRO, a multiplatform media company that claims to advocate for the perspectives of BiPOC. In this paper we ask: Does algorithmic gossip about injustice and bias enacted by human and non-human actors on a platform provide a basis for overcoming algorithmic oppression or is it inevitably incorporated into hegemonic strategies of platform governance and silencing? From an intersectional perspective, we acknowledge that, historically, the voices of BIPoC (Black, Indigenous and People of Color) and LGBTQ+ (Lesbian, Gay, Bisexual, Transgender and Queer) have been silenced, their suffering minimized or erased from public consciousness (D’Ignazio & Klein, 2020). With our research, we draw on Sophie Bishop's concept of algorithmic gossip:Algorithmic gossip refers to the communally informed strategising with regards to recommender algorithms, which is then communicated among creators and put to the task of engendering stable income and a spot under the algorithmic limelight within the precarious space of social media platforms. With this concept, Bishop (2019) acknowledges the technical expertise of content creators as well as the productive capabilities of gossip as a tool for unveiling experiences, assumptions and information about “how algorithms work” (1). We question the legitimacy of platform companies being able to have singular authority over issues of algorithmic visibility. In this paper, we have two core arguments: We argue that end users gather insights from the analytics on the creator studio of Instagram, YouTube and TikTok in order to reveal patterns of algorithmic injustice and bias. To substantiate this argument we first conduct a content analysis of the creator studio feature on Instagram, YouTube and TikTok, where we offer a reading of the interface of this feature and highlight a key similarity: data analytics are structured within the digital architecture of the platforms to ‘nudge’ creators towards allying their content production with the political and economic agenda of the platform company. We then refer back to our case study, involving content creators who published videos relevant to the Black Lives Matter protests only to then discover that their channels stopped growing and their views only declined. By manually collecting videos commenting on this injustice while using images of their own creator studio, and then investigating their impact on TikTok and mainstream media coverage, we posit that black creators use the creator studio as a tool to expose algorithmic injustice and to support the Black Lives Matter movement. Whereas content creators cannot directly intervene in the algorithmic architecture of the platforms they use, nor can they manipulate elements in the interface of the Creator Studio, they recontextualize the data they can access in order to weave counter-hegemonic narratives. We shed light on examples of this practice and explain how creators repurpose the analytics from the Creator Studio using algorithmic gossip. By demonstrating how end users collectively share information, we seek to extend academic understanding on algorithmic gossip, accentuating its potential as decolonial practice from below. In other words, gossip allows BiPOC to advocate for themselves, against algorithmic injustice and racial biases. However, our case does not end here. In the second part of this paper, we argue that platform companies undermine the collective struggle of BiPOC groups through self-governance. As shown above, TikTok was not held accountable for racial bias and for suppressing minoritized communities by external actors. Rather, the company was able to circumvent public outcry by launching an incubator program for black creatives. However, we argue that these incubator programs can be seen as attempts to silence the algorithmic gossip of BiPOC at the surface level. While establishing the program, the platforms do not effectively change the status quo under which BiPOC are oppressed within the platform. Moreover, whereas TikTok openly supports the entrepreneurialisation of BiPOC within the platform, they do not acknowledge that the creators who protested against TikTok’s “algorithms of oppression” (Noble, 2018) did not do so for the sole reason of building a brand or making a profit. In other words, TikTok’s response to the accusations of racial bias and the repression of the voices of BiPOC actively installs strategies to ignore the changes that minoritized groups demand of platform companies to transform oppressive structures and hence decolonize the algorithmic architecture of TikTok. Instead, they shift the conversation in such a way that the company stays in control of the changes that take place within the platform. They reserve the right to approve candidates for the incubator program, they maintain a highly profitable status quo, and ultimately they continue to colonize the voices and existence of BiPOC within TikTok. As a result, we contend that this case study legitimizes arguments in the discourse against the self-governance of platform companies. To ground our critique, we have synthesized literature dealing with oppressive algorithmic architectures (Umoja Noble 2018), the silencing and erasure of BiPOC (Nelson, 2016; D’Ignazio & Klein, 2020) and platform governance (Gorwa, 2019; Van Dijck, 2021). In conclusion, we follow data analytics from the TikTok Creator Studio as they become an instrument of protest in the hands of BiPOC, allowing content creators to fight for algorithmic justice from below. We then bear witness to the problematic mechanics of platform self-governance, offering a critique for TikTok’s attempt to practice social justice through an incubator program. We draw attention to the tensions that this case study introduces for the political economy of platforms and we urge the academic community to consider and question which forms of governance can lead to meaningful change in the arduous process of decolonizing algorithmically driven social media platforms. Thereby, in this paper we offer an alternative, intersectional perspective to platform data infrastructures and their social, economic and cultural impact on TikTok’s participatory culture. References Bishop, S. (2019). Managing visibility on YouTube through algorithmic gossip. New Media & Society, 21(11–12), 2589–2606. https://doi.org/10.1177/1461444819854731 Burgess, J., & Green, J. (2018). YouTube: Online video and participatory culture. John Wiley & Sons. D’Ignazio, C., & Klein, L. F. (2020). Data feminism. The MIT Press. Gorwa, R. (2019). What is platform governance? Information, Communication & Society, 22(6), 854–871. https://doi.org/10.1080/1369118X.2019.1573914 Introducing the TikTok for Black Creatives incubator program | TikTok Newsroom. (o. J.). Abgerufen 2. März 2022, von https://newsroom.tiktok.com/en-us/tiktok-for-black-creatives-incubator-program Nelson, A. (2016). The social life of DNA: Race, reparations, and reconciliation after the genome. Beacon Press. Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New York University Press. Van Dijck, J. (2021). Seeing the forest for the trees: Visualizing platformization and its governance. New Media & Society, 23(9), 2801–2819.