Selected Papers of #AoIR2024: The 25th Annual Conference of the Association of Internet Researchers Sheffield, UK / 30 Oct – 02 Nov 2024 Suggested Citation (APA): Lemos, A., Estima, W. (2024, October). Autonomy under surveillance: a failed experience on platform cooperativism in Brazil. Paper presented at AoIR2024: The 25th Annual Conference of the Association of Internet Researchers. Sheffield, UK: AoIR. Retrieved from http://spir.aoir.org. DISCIPLINED AUTONOMY: A FAILED PLATFORM-COOPERATIVISM EXPERIENCE IN BRAZIL André Lemos Faculty of Communication, UFBA Walmir Estima Faculty of Communication, UFBA Introduction Platformisation (Van Dijck et al., 2019) operates as a global data extraction and processing structure on proprietary platforms that have conquered central positions in international socioeconomic relations. With its well-established operating rules and power relations, the global platform ecosystem poses challenges and limitations to the platform cooperativism movement. Platform cooperativism is a movement that seeks to unite traditional cooperative values with digital communication technologies (Scholz, 2016). It aims to offer a counterpoint to platformisation in the struggle for autonomous and fairer working conditions concerning commercial platforms. The analysis of a failed partnership involving a rideshare driver cooperative, Coomappa, in the city of Araraquara, São Paulo, Brazil, the city's municipality and a private company (Bibi Mobilidade) provides an opportunity to discuss platform cooperativism and examine the manifestations of power in the Global South. The dynamics of platform capitalism (Srnicek, 2017), data colonialism (Couldry & Mejias, 2019), and surveillance capitalism (Zuboff, 2019) constituted the operating model of the experience that is established in the contradiction between the discourse of autonomy and the control of extractive data and resource relations through a disciplinary (economic, technical, organisational) power (Foucault, 1977). In the case under analysis, this power is what we name “disciplined autonomy”. A Failed Experience The experience failed after eight months of operation. Based on the premise that failure reveals sociotechnical networks (Appadurai & Alexander, 2020; Baker & Korolkova, 2021; Bowes & Korolkova, 2020; Lemos, 2023; Rettberg, 2023), we argue that infrastructural dependencies defined local actors’ scope of action from the outset, creating disciplined autonomy: the apparent self-governance of a resistance movement within boundaries set by disciplinary power. Coomappa was born from mobilising drivers in Araraquara to create alternatives to the low payments received for rides on Uber and 99 Pop apps. In feasibility studies, the cooperative decided not to develop its app and to invest in the Bibi Mob app franchise from Bibi Mobilidade. The city hall welcomed the project through the Public Incubator of Creative and Solidarity Economy. Its launch generated many celebratory headlines in Brazil's traditional news portals. Brazilian platform cooperativism initiatives are experimental (Grohmann, 2021, p.3). Method: Tracing power relationships through data relations The power of global infrastructural platforms is operating in the locatable action flows of the entities involved in the experience. Following the data, we locate globally redistributed local action from the continuous connections of the various stakeholders. We track the data produced by the experience to show how power was performed. The device is analysed based on official documents and website (terms of use and privacy policy), on the application interface (for users and drivers), on the analysis of Software Development Kits and Android application permissions (80% of Brazilians use the Android system), analysis of Araraquara City Law 10.161/23 and an in-depth interview with Coomappa's president. From the analysis, four major infrastructural platforms were involved in the process from the beginning: Alphabet, Meta, Apple and Amazon. Passenger and driver data are stored and processed on Amazon servers coupled with all application usage data. Alphabet (Google Maps) controls the geolocation data of registered users and drivers. Support and interpersonal communication activities are carried out by WhatsApp (Meta), installed on users' and drivers' smartphones, as well as Bibi Mobilidade attendants and Coomappa collaborators. Download and installation activities depend on Google and Apple stores. The identified SDKs show data flow toward Amazon, Meta, and Google. The permission to access user accounts in the passenger app, the only one with Facebook's data tracking, indicates that the data produced by Bibi Mob is shared with Meta for advertising purposes. Disciplined autonomy In the case of digital platforms, autonomy presupposes control of the platform's operation regarding its algorithmic logic, data usage, interfaces, organisation, and business model. Coomappa did not have self-determination over the Bibi Mob platform, which, in turn, was dependent on other infrastructural platforms such as Google, Amazon, Apple, and Meta. However, both the sectoral and the global infrastructural platforms seemed to give workers autonomy over their jobs, and that is why they welcomed the partnership, guided by better remuneration. However, before long, autonomy was revealed as partial, monitored, and controlled by mega-platforms that dictated operating costs by changing pricing and narrowing the remuneration possibilities, making the initiative unfeasible for drivers. The disciplinary power produces disciplined autonomy through the Bibi Mob app and affects the workers' cooperative. It operates in the locatable action flows of the entities creating the experience (drivers, passengers, app, city hall, cooperative, and private company), and it is materially detectable through the mapping of data flow and its management. Conclusion The "failure" does not refer to a digital platform that did not work, but quite the opposite, by an operational rationale established by global platforms that led to the bankruptcy of the sociotechnical arrangement. It indicates that the central element of the infrastructural platforms' power is giving relative autonomy to workers controlled by a business model that concentrates organisational power (tariffing, technologies, strategies, data flow surveillance). This research shows that platforms promote partial autonomy as a narrative to make partners docile by offering levels of flexibility that result, in the end, in poor working conditions, low remuneration, and dependence (Morales & Stecher, 2023; Wood et al., 2019; Kerssens & Van Dijck, 2022). The failure of the Araraquara experience is due to the exercise of the platform's disciplinary power applied to the partners as a disciplined autonomy. The franchised platform was a mediator of these processes by embedding data capture in its structure, incorporating infrastructural services from SDKs and inevitable infrastructural dependencies from Google Maps, App Stores, and Amazon Web Services. Bibi Mobilidade company had more control over the datafication process than Coomappa because it controlled monetisation from Facebook SDKs and Open Telemetry functionalities and used driver and user data to try to continue operations in the city without Coomappa. In a power pyramid, Coomappa would be below Bibi Mobility. The GAFAM is on top, orchestrating relationships. The power of platforms revealed by the failure of the Brazilian experiment occurred through a disciplined autonomy. This power ensures domination through control over data, imposition of standards, and infrastructural dependencies that affect the definition of business models and the field of action of those on the periphery of the structure. The concept of disciplined autonomy adds evident contradictions to the process of platformisation: unwitting reproduction by a resistance movement of the same power relations it seeks to counter. The main challenges posed to the platform cooperativism movement are in this tangle of relational power. Without a broad and efficient data sovereignty policy, the autonomy of cooperative movements will always be watched and at the mercy of disciplinary power – even if they achieve partial and temporary success. Platform cooperativism needs to be situated within a broader framework of public policies to offer alternatives to infrastructural platforms. In the Global South, policies for innovation and the development of alternative digital infrastructures are urgent. References Appadurai, A., & Alexander, N. (2020). Failure. Polity. Barker, T., & Korolkova, M. (Orgs.). (2022). Miscommunications: Errors, Mistakes, Media. Bloomsbury Academic. Couldry, N., & Mejias, U. A. (2019). Data colonialism: Rethinking big data's relation to the contemporary subject. Television & New Media, 20(4), 336–349. https://doi.org/10.1177/1527476418796632 Foucault, M. (1977). Discipline & Punish. New York: Vintage. Grohmann, R. (2021). Rider Platforms? Building Worker-Owned Experiences in Spain, France, and Brazil. South Atlantic Quarterly, 120(4), 839–852. doi:10.1215/00382876- 9443392 Kerssens, N., & van DIJCK, J. (2022). Governed by Edtech? Valuing pedagogical autonomy in a platform society. Harvard Educational Review, 92(2), 284-303 Korolkova, M., & Bowes, S. (2020). Mistake as method: Towards an epistemology of errors in creative practice and research. European Journal of Media Studies, 9(2), 139– 157. https://necsus-ejms.org/mistake-as-method-towards-an-epistemology-of-errors-in- creative-practice-and-research/ Latour, B. (2005). Reassembling the social: An introduction to actor-network-theory. Oxford University Press. Lemos, A (2023). Errores en la cultura digital. In Carlon, M. Lo contemporáneo: indagaciones sobre el cambio de época en/desde América Latina. Universidad de Buenos Aires. Instituto de Investigaciones Gino Germani. pp. 65–90. Morales, K., & Stecher, A. (2023). Platform capitalism and neo-normative control: "autonomy" as a digital platform control strategy in neoliberal Chile. New Technology, Work and Employment, 38(2), 230-251. Rettberg, J. W. (2022). Algorithmic failure as a humanities methodology: Machine learning's mispredictions identify rich cases for qualitative analysis. Big Data & Society, 9(2), 205395172211312. https://doi.org/10.1177/20539517221131290 Scholz, T. (2016). Platform cooperativism. Challenging the corporate sharing economy. New York: Rosa Luxemburg Foundation. Srnicek, N. (2017). Platform capitalism. Cambridge: Polity http://catalogoiigg.sociales.uba.ar/cgi-bin/koha/opac-retrieve-file.pl?id=792f22d829f864d0c4288885abbc188f http://catalogoiigg.sociales.uba.ar/cgi-bin/koha/opac-retrieve-file.pl?id=792f22d829f864d0c4288885abbc188f Van Dijck, J. & Nieborg, D. & Poell, T. (2019). Reframing platform power. Internet Policy Review, 8(2). https://doi.org/10.14763/2019.2.1414 Wood, A. J., Graham, M., Lehdonvirta, V., & Hjorth, I. (2019). Good gig, bad gig: autonomy and algorithmic control in the global gig economy. Work, employment and society, 33(1), 56-75. Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. Public Affairs. https://doi.org/10.14763/2019.2.1414