Regulators are increasingly debating how to regulate access to and use of social media platforms by children and adolescents. However, emerging evidence calls into question that these regulations achieve the desired ends raising the question of why it seems so difficult to regulate how children and adolescents engage with social media platforms. Research in organization studies and information systems (IS) has oftentimes adopted institutional theory to study regulation of companies and technologies. One way to use this theory to study regulation of companies is a view of „ institutional politics“ where institutions such as the state have systemic power to regulate actors such as private companies. In our conceptual paper, we critically engage this view,  and extend it to predict why regulation of platform companies is hard when it comes to use of social media platforms by kids and adolescents. Specifically, we argue that in addition to classical views of institutional politics where states regulate companies, regulation of platform companies is a „reverse process of institutional politics“ where platform power is a form of systemic power that shapes institutions. We offer testable propositions revolving around how AI models perform statistics („actuarial practices“) lets platform power arise and increase relative to power of the state. In doing so, we extend how we as field think about institutional politics, and offer broader implications of institutional theory as well. 

Key readings: 

Lawrence, T. B. (2008). Power, institutions and organizations. In R. Greenwood, C. Oliver, K. Sahlin, & R. Suddaby (Eds.), *The SAGE handbook of organizational institutionalism* (pp. 170–197). SAGE Publications. 

This thesis focuses on one particular form of implicit rhetoric, namely dogwhistles. A growing interest in contemporary political discourse has led to an increasing body of research on implicit communication and, more specifically, on dogwhistle rhetoric. In its literal sense, a dogwhistle is a whistle that emits a high-frequency sound audible only to certain animals while remaining imperceptible to humans. In political discourse, the term refers metaphorically to coded messages that allow speakers to convey different meanings to different audiences through the same utterance. Thus, speakers can appeal to a target audience while maintaining plausible deniability before a broader public. Against this backdrop, this thesis investigates the use of dogwhistles in a corpus of social media posts published by European politicians between 2020 and 2024. More specifically, it analyses how politicians from Spain, the United Kingdom, and Germany employ dogwhistle rhetoric on X (formerly Twitter), providing a comparative account of its frequency and discursive realisation across the three contexts.   

Key readings:  

Saul, Jennifer. (2018). Dogwhistles, Political Manipulation, and Philosophy of Language. In Daniel Fogal, Daniel W. Harris, and Matt Moss (Eds.), New Work on Speech Acts (pp. 360–383). Oxford. https://philarchive.org/archive/SAUDPM (Open Access Draft Version) 

Henderson, R., & McCready, E. (2021). Dogwhistles: Persona and Ideology. In N. Dreier, C. Kwon, T. Darnell, & J. Starr (Eds), Proceedings of SALT (pp. 703–719). https://doi.org/10.3765/ywwkd646  

Organisations increasingly rely on data repurposing, that is, reshaping existing data for tasks other than those originally planned. Yet, while data repurposing can bring substantial value, data created for one purpose rarely meet the requirements of new goals. At the same time, data quality (DQ) – which depends on the context of use – cannot be assessed against pre-determined criteria when requirements are unknown in advance.  

Adopting a practice lens, I examine how data repurposing is pursued in practice through twenty-five semi-structured interviews with data professionals, analysed inductively following the Gioia methodology. Findings show that data repurposing is an iterative process evolving along three interrelated areas – working towards task, working towards data, and working towards data-task compromise. I argue that data repurposing involves both flexible and stable practices, allowing it to be conceptualised as structured flexibility. This study extends Parsons et al.’s (2026) data repurposing framework and the DQ debate into the realm of data work, offering practical insights for organisations repurposing data. 

Key readings:  

Parmiggiani, E., Østerlie, T., & Almklov, P. G. (2022). In the Backrooms of Data Science. Journal of the Association for Information Systems, 23(1), 139–164. [Link] 

Parsons, J., Lukyanenko, R., Greenwood, B. N., & Cooper, C. B. (2026). Understanding and Improving Data Repurposing. MIS Quarterly. [Link] 

Key readings:  

Lilie Chouliaraki (2015) Digital witnessing in conflict zones: the politics of remediation, Information, Communication & Society, 18:11, 1362-1377, DOI: 10.1080/1369118X.2015.1070890 

Dan, V. (2017). Why so few integrative framing analyses? In Integrative framing analysis: Framing health through words and visuals (Chapter 3). Routledge. https://doi.org/10.4324/9781315171456 

How should the law respond when AI systems generate false and defamatory information about real people? Using AI hallucinations as a case study, this talk explores how the EU regulatory architecture, comprising i.a. the EU AI Act, GDPR, DSA, and national liability law, addresses emerging AI harms, and what this reveals about the EU’s broader approach to balancing innovation and effective regulation. 

Key readings:  

Lilian Edwards, Reuben Binns: Reputation Management in the ChatGPT Era, https://arxiv.org/abs/2412.06356  

Key readings:  

Meyer, L. 2026. “ Regulatory Power Beyond the Brussels Effect—How Data Localization Requirements Restructure Global Cloud-Infrastructures.” Contemporary European Politics 4: e70041. https://doi.org/10.1002/cep4.70041 .