2026-10-06T08:56:48Z https://api.figshare.com/v2/oai

oai:figshare.com:article/32805107 2026-10-01T16:18:19Z portal_1172 item_type_8 month_year_10_2026

Dynamics for confidence: safety, security and legal compliance of algorithmic decision-making processes David Mark (171458) PUREID: 660803741 Algorithmic decision-making public sector WPR safety security legal compliance comparative regulation EU–US–China This thesis investigates the governance of public sector algorithmic decision-making (ADM). Governments worldwide are increasingly deploying algorithmic systems to inform or determine decisions with profound legal and social consequences. This expansion has in turn prompted accelerating regulatory activity and intensified geopolitical contestations.<br> <br>In this context, three paradigmatic jurisdictions, the European Union (EU), the United States (US), and the People’s Republic of China (PRC), have emerged as regulatory benchmarks for the governance of ADM on the international stage. Each has its own distinctive initiatives and claims as to how such technology should be regulated, and how to ensure ‘safety,’ ‘security,’ ‘legal compliance,’ and ultimately ‘trustworthy’ deployment of ADM systems.<br><br>This research examines the regulatory frameworks emerging from these three actors, adopting a problematisation-oriented approach grounded in Carol Bacchi’s What’s the Problem Represented to Be? (WPR) framework. This is applied through the intersecting lenses of safety, security, and legal compliance, to a selected tri-jurisdictional corpus of prescriptive regulatory texts. The resulting comparative analysis demonstrates that despite a common vocabulary and notably overlapping aspirations, the three jurisdictions problematise the risks associated with ADM in markedly different ways. In doing so, these divergent problem representations embed distinct assumptions, foreground particular rationalities, and generate silences. Ultimately, they produce heterogeneous, and at times conflicting, regulatory approaches, leaving the governance landscape uneven and contested. This research is situated within this contested landscape. 2026-10-01T16:18:19Z Text Thesis 10.17034/32805107.v1 https://figshare.com/articles/thesis/Dynamics_for_confidence_safety_security_and_legal_compliance_of_algorithmic_decision-making_processes/32805107 All Rights Reserved