This project critically considers the rapid diffusion of so-called generative AI: that is, machine learning technologies that simulate human languages, communication, arts, and cultures through the statistical modelling of vast troves of scraped internet data.

As these commercial technologies aim to simulate and mediate human expression and creativity at an unprecedented scale, this research supports interdisciplinary alliances between researchers, students and community partners. Our goal is not only to critique these fast-developing technologies, but also to envision machine learning systems that are safe, accountable, inclusive and receptive to many voices.

This collaboration is inspired by the work of the Design Justice Network, a hub for people committed to embodying and practicing the Design Justice Network Principles. The project facilitates research that combines interdisciplinary critique, best practices from data science and digital humanities, and collaborative research that strives to centre people and cultures that have been marginalized by design processes. A key outcome is the Design Justice AI Institute, hosted by University of Pretoria in 2024 and a digital repository of findings and research templates based on humanistic study of the simulation of human language, communication, art and culture by AI software.