Knowers and others: Indigenous representation in search (and answer) engines

Aboriginal knowers and knowledges have historically been sidelined in accounts of ‘authoritative’ information about us across a range of knowledge institutions. This epistemic exclusion catalysed struggles to self-represent and counter dominant images and stories, critically reread histories, and reframe the way issues that concern us are understood and discussed. Google is a powerful epistemic actor. Through design choices, Google can facilitate self-representation or reinforce the authority claimed by others to represent Aboriginal peoples. The integration of large language models into search systems is reshaping how users encounter and engage with information online, raising urgent questions about online visibility and representation. What representations are prioritised in Google's direct answers? Which knowers are treated as authoritative? What does this mean for Indigenous representation? This paper contributes to the empirical basis for reflection on these critical questions with a case study of Google AI overviews. Using thematic and close textual analysis of a corpus of question-direct answer pairs, the study offers insights into how AI overviews reinforce or accommodate challenges to dominant discourses, and more generally how ubiquitous, everyday applications of AI are affecting Aboriginal self-representation. The article argues that the shift to AI-generated search results severs links between knowers, knowledge and context in different and subtle ways, with serious implications for Indigenous representation, rereading and reframing.