New Technologies from a legal standpoint – Sciences Po Paris / March 2019 Article Review: K. Yeung, « Algorithmic regulation: a critical interrogation », Regulation & Governance, 2018. Written by Ricardo Zapata L. Silicon Valley entrepreneur Tim O’Reilly (2013) was the first one to propose the concept of algorithmic regulation as a practical method for improving efficiency in public policy delivery. In a provocative 1 article, he highlighted that “regulation (…) should be regarded in much the same way that programmers regard their code and algorithms” (p. 291), with a constantly updated toolset focused on achieving pre-specified outcomes. But O’Reilly never defines algorithmic regulation, he just points four shared features of algorithmically regulated systems 2 and how they are already applied in different economic sectors. British scholar Karen Yeung departs from this to, first, build a working definition and a taxonomy of algorithmic regulation (AR), and second, critically interrogate the concept and its implications. Her previous academic background in understanding regulation by design and its ethical, political, legal and social implications frames the article's purpose, which makes part of a series of other critical interrogations on contemporary technological trends like artificial intelligence, the blockchain, machine learning, and bid data. She finds that the critical analysis of AR is relevant due to its capacity to generate a new system of social ordering and asymmetric power relationships. Thus, she concludes, the political and regulatory debate is much more than transparency and accountability. Nevertheless, it could be disappointing for practitioners that must take urgent decisions, as the debate is left open. AR, according to Yeung, is a concept related to the realm of big data and machine learning. The reviewed article aims to understand the underlying ethical, legal, social and political logics behind algorithmic mediated settings. In the first section, Yeung develops a definition and a taxonomy of AR; in the second, she critically interrogates it from various academic stands (regulatory governance and public administration, legal scholarship, surveillance studies, and critical data studies) and maps the present debates and positions. Yeung works on the first consistent definition of AR. Although other authors had previously worked on similar concepts, her main contribution is to provide a new analytical category that fits contemporary technological applications. She defines AR “as decision-making systems that regulate a domain of activity in order to manage risk or alter behavior through continual computational generation of knowledge from data emitted and directly collected (in real time on a continuous basis) from numerous dynamic components pertaining to the regulated environment in order to identify and, if necessary, automatically refine (or prompt refinement of) the system’s operations to attain a prespecified goal.” (p. 507). 1 Both for the enthusiasm it aroused, and the critical questions it sparked. 2 “(i) a deep understanding of the desired outcome. (ii) real-time measurement to determine if that outcome is being achieved, (iii) algorithms (i.e. a set of rules) that make adjustments based on new data, and (iv) periodic, deeper analysis of whether the algorithms themselves are correct and performing as expected (O’Reilly 2013)” (p. 506). Graphical interpretation of Algorithmic Regulation’s definition by Karen Yeung (2018a).