AI and I—Distributed Authorships in the Swedish Upper Secondary Writing Classroom

Abstract This paper examines upper secondary students’ writing with Generative Artificial Intelligence (GenAI) through a post-structural lens. The aim is to explore how authorship can be understood and described in the age of AI. The empirical material consists of three students’ recorded writing sessions and complementary interviews. Through rhizomatic mapping, the recordings are analyzed with Deleuze and Guattari’s notions of assemblage, affect and becoming, further extended through the concept of distributed authorship. The analyses map three dynamic assemblages in which writing and authorship are redistributed across heterogeneous relations. One assemblage foregrounds AI as the primary authorial agent; another enacts reciprocal human-AI co-writing; and a third positions AI as a threat to writing, creativity, and the climate. Writing with GenAI unfolds as a complex relational practice and authorship emerges as distributed between humans and non-humans. This study raises important questions for education to consider in writing instruction and in developing policies for writing with AI, such as questions of equity and ethics.