This manuscript is a preprint and has not undergone peer review. It examines how the institutional integration of generative AI, particularly large language models, reconfigures the procedural architecture of responsibility in hybrid decision-making environments. Drawing on responsible innovation theory, the article argues that the central transformation is procedural: epistemic labour is redistributed, justificatory narratives are co-produced within hybrid infrastructures, and the pathways through which accountability is assigned become reconfigured.
Han et al. (Wed,) studied this question.