As we implement AI technologies across Monash, it is important to recognise the specific risks they pose for Aboriginal and Torres Strait Islander peoples. Many large language models are built from material largely created by non‑Indigenous perspectives. As a result, they can reproduce harmful narratives, misrepresent culture and misuse Indigenous Cultural and Intellectual Property (ICIP). In some cases, Indigenous languages, knowledges and cultural expressions have been incorporated into AI datasets without consent, repeating long‑standing patterns of extraction and misappropriation.
Current AI systems do not adequately reflect the depth, diversity, or cultural authority of Indigenous knowledge systems in Australia. Because most of the material these models draw from has been written about Indigenous peoples, not by Indigenous peoples, AI outputs can amplify biased, incomplete, or inaccurate representations. These risks include the perpetuation of unsafe narratives, the distortion of cultural meaning, and the inappropriate reproduction, ownership or circulation of Indigenous knowledges, artworks, stories and other forms of ICIP.
Beyond technical limitations, these issues raise deeper questions of cultural authority, ethics and benefit. Indigenous knowledges are not simply information to be collected or reproduced. They exist within relationships, responsibilities and community‑held protocols. Without careful governance, AI technologies may unintentionally undermine cultural authority, contribute to cultural appropriation or erode years of progress made in universities to strengthen respect for Indigenous authorship and knowledge stewardship.
Monash has made significant progress in recent years in recognising and embedding Indigenous knowledges and perspectives across teaching, research and engagement. It is important that the introduction of AI technologies does not undermine that work.
For this reason, AI tools should not currently be used when engaging with Aboriginal or Torres Strait Islander knowledges, cultures or data.
We are working closely with Indigenous colleagues to better understand these risks and to explore how Indigenous knowledges can be appropriately represented and governed within emerging AI technologies. This work will guide how AI is used across the University in ways that uphold cultural authority, respect Indigenous data sovereignty and align with Monash values.
