The effects of artificial intelligence on institutional pluralism. Case study on AIS within the French judicial field

How do AIS affect the institutional pluralism of highly structured institutional fields? This paper addresses this question by examining the deployment of AIS in the French judicial field. Adopting a strong process-oriented perspective (Reinecke & Lawrence, 2023) combined with the theory of institutional logics (Thornton et al., 2012; Fak et al., 2020), we view the deployment of AIS not as a rupture, but as an incremental and endogenous transformation. We examine how fields absorb the integration of new technology. Based on a qualitative case study comprising 24 semi-structured interviews, two nonparticipant observations, and an analysis of 80 documents, we present three key findings. Firstly, we identify a four-stage stabilisation cycle (infusion, impetus, acceleration and absorption) through which the legal sector gradually adopts the technology. Secondly, we demonstrate that AIS is more than just a tool. We are witnessing the stabilisation of a new institutional logic (the algorithmic logic) driven by nine mechanisms involved in its framing, integration, and finally, anchoring. Thirdly, we show that this algorithmic logic affects institutional pluralism in three ways: it feeds off dominant logics to establish its legitimacy; it reinforces managerial logic by amplifying its bases of attention and strategy; and it gradually weakens bureaucratic logic first, and then more insidiously, legal logic, by calling into question their professional identities, sources of authority, and original metaphors. Technology does not directly challenge existing logics, but rather gradually renders them obsolete, ultimately reducing their influence within institutional pluralism.