ABSTRACT The integration of Generative AI into artistic practices is not only expanding the possibilities for creative expression but also challenging foundational concepts such as agency and authorship. Combining Agential Realism and Speech Act Theory, this paper presents a case study involving three internationally recognised artists with extensive experience using text and image-based models. It combines in-depth interviews with critical readings of selected artworks to explore how generative AI Art emerges from the intra-actions between artists, data, and algorithms. Our contribution unfolds along two complementary directions. Firstly, we engage with epistemological debates by conceptualising agency and authorship as relational and distributed phenomena. Secondly, we challenge the notion of AI as a mere tool and advocate for design approaches that foreground shared authorship. By focusing on studio-level practices, this paper offers situated insights into artist–AI interaction, expands our understanding of de-individualised knowledge production, and points toward systems that support responsible co-creation.

The design of robust and trustworthy Generative AI (GenAI) requires a deep understanding of the agencies emerging from human interactions with them. To contribute to this goal, we retrospectively studied an art project involving a visual artist, a computer scientist, an artistic director, and a generative model (GPT-2). The model was fine-tuned with trip reports describing the experience of eating psychedelic mushrooms. Building on agential realism, we analysed the co-performance between the artist and the model as their agency moved along the choreographer-performer continuum. Results reveal ontological surprises, leading to the proposal of entangled authorship to de-individualise the production of knowledge from a More Than Human perspective. The paper illustrates how art can expose different forms of relationships, challenging the idea of GenAI as just a tool that simplifies or replaces human labour. We conclude by emphasising the transformational potential of GenAI for novel modes of engagement between humans and machines.