Melbourne school cracks down on AI cheating in Year 12 exams

An all-boys school in Melbourne's outer east has taken disciplinary action against several Year 12 students caught using artificial intelligence tools to cheat during formal examinations, raising fresh concerns about academic integrity in Victoria's secondary education system.

Mazenod College in Mulgrave discovered the breach during an oral English exam, with principal Paul Shannon confirming that students had deployed AI tools to gain unfair advantage in the assessment task. While the school declined to specify the number of students involved, all were spoken to following a thorough investigation, and their marks on the affected assessment were subsequently reduced.

Integrity measures under strain

"Written and oral assessment tasks, especially exams, are designed to measure a student's own understanding and independent thinking," Mr Shannon said in a statement to ABC News. "While the use of AI tools is a growing challenge within all schools, they have no place in formal examinations."

The incident comes as schools across Australia grapple with the rapid proliferation of generative AI tools, which have become increasingly accessible to students and capable of producing sophisticated academic work. The discovery at Mazenod College suggests that traditional examination protocols may be inadequate in preventing such breaches.

Victoria's regulatory framework under pressure

The Victorian Curriculum and Assessment Authority (VCAA), which oversees secondary education standards in the state, has explicit rules stating that unauthorised use of AI may constitute a breach of academic integrity. However, the regulatory body has faced criticism for lacking comprehensive guidance on how schools should detect and prevent AI-assisted cheating.

The VCAA's current framework allows limited, disclosed use of AI in some learning contexts, but categorically prohibits its use in formal examinations without explicit permission. Mazenod College's response—reducing the affected students' marks—appears consistent with standard academic misconduct protocols, though questions remain about whether such penalties adequately deter future breaches.

Broader concerns for education sector

The Mulgrave incident reflects a widening challenge confronting Australian secondary schools. As ChatGPT and similar tools become ubiquitous among young people, educators face mounting pressure to redesign assessment methods that genuinely test student knowledge rather than capacity to prompt AI systems effectively.

Education experts have suggested that schools must adapt by:

• Implementing more authentic, real-time assessment tasks that cannot easily be completed by AI
• Developing clearer policies distinguishing permitted from prohibited AI use
• Providing staff training in detecting AI-generated work
• Engaging students in digital citizenship education around ethical technology use

The matter has been submitted to the VCAA through the school's formal processes, ensuring the incident is recorded within Victoria's education regulatory system.

Source: ABC News