AI is wrecking higher education but the system was already broken

The rise of generative AI in Australian universities exposes a structural crisis in the sector. According to antiquarian booksellers, AI companies are bulk-buying old books to train their models through a process called "destructive scanning." Lecturers now face suspiciously flawless student papers, struggling to distinguish human work from algorithm-generated text.

A 2024 study found 40% of Australian students admitted to using AI for cheating, while 2025 surveys show nearly 80% of Australian and two-thirds of American students rely on AI, though often without specifying for what tasks. Some universities, like the University of Adelaide or RMIT, even encourage staff to automate grading, under human supervision. Academics denounce widespread fraud, while others, like Western Sydney University’s Cath Ellis, urge students not to outsource their thinking.

This AI dependency stems from decades of neoliberal reforms that hollowed out higher education’s pedagogical core. Initially adopted for mundane tasks, AI tools now reshape core academic processes. Yet recent cases reveal even official university statements may be partially AI-generated, highlighting the scale of the issue.