The rapid integration of Generative Artificial Intelligence (GenAI) in education constitutes a critical dimension of the poly-crisis facing higher education institutions, threatening trust in the value of its knowledge production function. Dominant institutional response to GenAI has been defensive and arborescent, rooted in surveillance, bans, or a retreat to invigilated examinations to protect the status quo of credentialing (Tsao, 2025). This paper argues that such attempts to out-design or police AI are futile, and defends a transactional and inherently neoliberal view of education that the technology can already simulate with increasing efficiency.

This submission reinvigorates pivots toward affirmative rhizomatic pedagogies, grounded in the philosophy of Deleuze and Guattari (1987). The concept of the rhizome aims to rethink teaching and learning in a post-digital, AI-saturated contemporary and how pedagogical and assessment practices could change when knowing is about becoming. Furthermore, this paper positions the rhizome as a necessary decolonial intervention, which has no centre, resists hierarchy, and values the periphery. The concept challenges the homogenising force of algorithmic bias that privileges epistemological colonisation, enforcing a statistical consensus derived primarily from Global North data and English-centric corpora, and the centralisation of epistemic power among technological elites.

Contrasting arborescent structures (hierarchy, linearity, and tracing) against rhizomatic networks (non-linearity, heterogeneity, and mapping), the paper posits that while GenAI outputs appear rhizomatic, the underlying architecture behind transformer or diffusion technology is fundamentally arborescent because they operate through probabilistic adjacency, error minimisation, and consensus-seeking, thus functioning as powerful engines of reproduction rather than novelty.

This research responds to the global urgency of redefining human learning, aligning with the European Union’s Digital Education Action Plan which prioritises critical digital literacy and human agency. The paper asks: How can we design learning that works in productive tension alongside AI? By applying the six principles of the rhizome—connection, heterogeneity, multiplicity, asignifying rupture, cartography, and decalcomania—this proposal outlines a transdisciplinary nexus (Gibbs, 2026) for education that secures the future of academia by cultivating “nomadic” learners capable of traversing the uncertain terrains that AI cannot map: the embodied, the situated, and the semantic.

This paper uses Deleuze and Guattari’s geo-philosophy and conceptual methodology through a “diffractive reading” to interrogate learning and architecture of GenAI. The inquiry proceeds in two analytical stages. First, it performs an ontological critique of GenAI capabilities and identifies the asymmetries between algorithmic processing and human learning. It argues that AI operates through tracing (reproducing deep structures and statistical patterns) and probabilistic adjacency (connecting tokens based on likelihood). In contrast, rhizomatic human learning relies on mapping (constructing new realities) and semantic novelty. Second, the paper operationalises these theoretical distinctions into pedagogical design principles. It maps the six principles of the rhizome against specific assessment challenges posed by AI. This involves a synthesis of recent critical educational technology scholarship (Selwyn, 2024; Gašević et al., 2023) and post-digital learning theories (Fawns, 2022). The conceptual framework critiques current assessment practices and speculates on and prototypes new rhizomatic approaches designed to be AI-resilient not by exclusion, but by requiring inputs that AI cannot access—specifically embodied experience, material heterogeneity, and the “n-1” dimensions of the learning process.

The paper concludes that higher education must move more towards three specific pedagogical pivots. The first and most apparent is the shift from assessment of static products to the “n-1” dimensions of the process. Teachers can evaluate the genealogy of ideas, including concept sketches, prompt logs, discarded drafts, and the iterative dialogues between student intent and AI output. Second, the shift from semiotics to heterogeneity; since AI processes representations rather than reality, rhizomatic assessment must anchor learning in heterogeneous assemblages by connecting text to the material world, embodied experience, and localised contexts where human “flesh” (Merleau-Ponty, 1968) retains epistemic purchase. Third, a movement from tracing to mapping, with the teacher assuming the role of cartographer-guides, supporting students taking AI outputs as tracings (consensus baselines) that must be returned to the map for critique and reterritorialisation. Ultimately, this work suggests that the poly-crisis that includes AI is an opportunity to shed archaic, transactional educational models. By (re)centring rhizomatic principles, universities can foster learning that is not about replicating established hierarchies, but about navigating rupture and uncertainty, ensuring that universities remain spaces of pluralistic, democratic inquiry rather than echo chambers of automated consensus.

Deleuze, G., & Guattari, F. (1987). A thousand plateaus: Capitalism and schizophrenia. University of Minnesota Press. Fawns, T. (2022). An Entangled Pedagogy: Looking Beyond the Pedagogy—Technology Dichotomy. Postdigital Science and Education, 4(3), 711–728. https://doi.org/10.1007/s42438-022-00302-7 Gašević, D., Siemens, G., & Sadiq, S. (2023). Empowering learners for the age of artificial intelligence. Computers and Education: Artificial Intelligence, 4, 100130. https://doi.org/10.1016/j.caeai.2023.100130 Gibbs, P. (2026). Transdisciplinary emergence: Public universities seeking to provide common good. In G. Kochhar-Lindgren, W. Sims-Schouten, & J. Tsao (Eds), Transdisciplinary Experiments. UCL Press. Merleau-Ponty, M. (1968). The visible and the invisible. Northwestern University Press. Selwyn, N. (2024). On the Limits of Artificial Intelligence (AI) in Education. Nordisk Tidsskrift for Pedagogikk Og Kritikk, 10(1). https://doi.org/10.23865/ntpk.v10.6062 Tsao, J. (2025). Trajectories of AI policy in higher education: Interpretations, discourses, and enactments of students and teachers. Computers and Education: Artificial Intelligence, 9, 100496. https://doi.org/10.1016/j.caeai.2025.100496