Sydney academic used AI to write article cautioning against AI use
Sydney academic used AI to write article cautioning against AI use

A senior academic at Western Sydney University used artificial intelligence to write an opinion piece that urged students not to rely on AI to 'cut corners'. The article, published in the Sydney Morning Herald last month, has since been removed from the newspaper's website.

Professor Cath Ellis, the university's pro vice-chancellor for quality and integrity, wrote in response to an earlier piece by academic Kylie Moore-Gilbert. Moore-Gilbert had advised her stepdaughter to reconsider university, arguing that students could outsource learning to AI. Ellis countered that the 'AI problem is real' but encouraged students to attend university and 'do the work'.

However, the column was flagged as AI-generated by the detection service Pangram. When questioned by Guardian Australia, the university confirmed Ellis had used AI in writing the piece. A spokesperson said Ellis uploaded 40,000 words of her own material into a Copilot Large Language Model, which summarised her knowledge and provided prompts for early drafts.

The university defended the use of AI as 'sophisticated and appropriate', stating that detection tools cannot judge appropriateness. The article did not disclose the use of generative AI, which violates the editorial policy of Nine, the Sydney Morning Herald's parent company. Nine's policy allows AI for research and ideas but not for writing stories, and any AI-generated material must be labelled.

Nine did not respond to questions, but editor Jordan Baker later said the article did not meet editorial guidelines and was removed. 'The Herald was not informed of the use of AI... Clearly this is unacceptable,' Baker said. The incident adds to a growing list of media outlets publishing undisclosed AI content, including Crikey and the New York Times.