Presentation Schedule


Presenter Registration Banner 5

Ghost Logics: Resistance, Transparency, and Liberation in Generative AI Outputs (101072)

Session Information: Culture, Inter/Multiculturalism & Language
Session Chair: Akiko Kano

Monday, 5 January 2026 15:05
Session: Session 3 (Parallel)
Room: Ala Moana Hotel: Plumeria Room
Presentation Type: Oral Presentation

All presentation times are UTC-10 (Pacific/Honolulu)

This paper aims to demonstrate how genAI misrepresents Indigenous identities by drawing upon colonialist tropes and historical stereotypes. The term phantasm is used to create critical distance by emphasising the spectral quality of generated identities, and system’s inability, in most cases, to convey the lived realities of Indigenous groups. As stochastic systems, genAI reassembles fragmented cultural markers into synthetic composites that may appear coherent on the surface. However, without critical AI literacy to counter automation bias, educators and students risk mistaking such outputs for truth, perpetuating societal bias. Because datasets include pornographic materials, representational risk can include explicit and inappropriate content, particularly when prompts combine gender with non-white identity markers, for example ‘a Māori woman’. These tendencies underscore the urgent need for critical strategies to ensure appropriate, equitable use in education contexts. Informed by my positionality as a Māori artist, author, educator and researcher, I present a scaffolded, classroom-ready methodology that treats Māori identity as a site of epistemic resistance. Centring personal cosmologies, this counterfactual strategy disrupts representational defaults while building competency in the critical evaluation of AI outputs. Two linked practices shape the approach: system reflexion (surfacing the model’s compositional habits and sources) and recursive dialogue (reusing and inverting prompts to track coherence and bias persistence). Framed as Ghost Logics, the methodology sequences cultural sensitivity, transparency, and equitable outcomes—positioning the AI educator-user as a guide to learners in chasing, reshaping, and dismantling harmful representations.

Authors:
Pamela Morrow, Waipapa Taumata Rau University of Auckland, New Zealand


About the Presenter(s)
Pamela Morrow is a Communications Studies PhD candidate with Creative Practice at Waipapa Taumata Rau University of Auckland. Her theoretical work critiques the Māori phantasmagoria conjured by generative-AI.

See this presentation on the full scheduleMonday Schedule



Conference Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Presentation

Posted by James Alexander Gordon

Last updated: 2023-02-23 23:45:00