The Rise of AI Cheating in Australian Higher Education
Australian universities are grappling with an unprecedented wave of artificial intelligence (AI)-assisted academic misconduct, prompting a nationwide push for stricter enforcement measures. Generative AI tools like ChatGPT have revolutionized learning but also created new avenues for cheating, with surveys indicating that up to 83 percent of students use AI in their studies and 40 percent admit to using it improperly on assessments. Academics report that 80 to 90 percent of students in some courses rely on AI for essays, quizzes, and even research papers, undermining the integrity of degrees earned at institutions like the University of Sydney, Monash University, and Deakin University. This crisis has led to a coordinated crackdown, with universities mandating in-person assessments to verify genuine student work.
The shift comes amid growing evidence that remote, take-home exams—accelerated by the COVID-19 pandemic—are no longer viable. Low lecture attendance, such as just 7 percent at the University of Western Australia, exacerbates the issue, as students outsource critical thinking to AI without developing core skills. As one senior academic noted, "university administrations are failing to crack down due to denial, inertia, and helplessness."
Major Measures: Bringing Students Back to Campus
In response, universities across Australia are rolling out robust policies requiring at least 50 percent of assessments to involve direct observation. This includes supervised pen-and-paper exams, even on weekends, oral defenses (vivas), in-class presentations, and practical lab work where students must explain their processes step-by-step. Former Monash University Chancellor Dr. Alan Finkel and ex-Murdoch University Chancellor Terry Budge advocate for 70 percent of final marks to come from invigilated on-campus tasks to restore credibility.
- Supervised Exams: Traditional timed tests under proctoring to eliminate AI use entirely.
- Viva Voce Defenses: Students verbally justify written work, revealing gaps in understanding.
- Process-Based Tasks: Submission of drafts, prompts used, and iteration logs to demonstrate authentic effort.
- Relational Assessments: Personalized questions or real-time responses tailored to individual progress.
Institutions like Curtin University are disabling unreliable AI detection features in tools like Turnitin from January 2026, classifying AI misuse as contract cheating instead. The University of Sydney employs a "two-lane" policy: AI permitted in open assessments if cited, banned in closed ones, with students detailing tools and prompts used.
Spotlight on the Australian Catholic University Scandal
The push for stricter measures gained urgency following the Australian Catholic University (ACU) debacle in 2024. ACU's use of Turnitin's AI detector flagged nearly 6,000 misconduct cases across nine campuses—90 percent AI-related—but many were false positives, with students wrongly accused based solely on algorithmic flags. Affected students endured months-long investigations, withheld results on transcripts, and demands for evidence like search histories, derailing job applications and causing immense stress. One nursing student considered quitting her career due to the ordeal.
ACU Deputy Vice-Chancellor Tania Broadley admitted figures were "substantially overstated," scrapped the tool in March 2025, and implemented staff training on ethical AI. Turnitin itself warns its reports "may not always be accurate," recommending human oversight.ABC News Report This case highlights the pitfalls of over-relying on detection software, now flagged in over a dozen Australian universities.
Student Leaders' Backlash: Unfair and Punitive?
The National Union of Students (NUS) has voiced strong opposition, labeling the crackdown "unfair" and citing ACU's false accusations as evidence of flawed enforcement. Students argue that mandating campus returns—even weekends—ignores accessibility for regional, disabled, or working learners, exacerbating inequities. "These changes punish everyone for the actions of a few," an NUS spokesperson stated, calling for AI literacy programs over surveillance.
On platforms like Reddit, students decry the shift: "AI detectors don't work... creating false accusations." Concerns include mental health impacts from invigilation stress and the irony of punishing AI misuse while universities partner with AI firms.
Academics Sound the Alarm: An Epidemic of Cheating
Senior lecturers estimate 80 percent of assignments involve AI, with some courses seeing 90 percent "fullbore cheating." Humanities tutors report over half their first assignments flagged, a sharp rise from 2023. The Tertiary Education Quality and Standards Agency (TEQSA) warns of evolving risks, urging centralized investigations.
| Statistic | Value | Source |
|---|---|---|
| Students using AI daily | 44% | Survey |
| Admit to cheating with AI | 40% | AcademicJobs |
| AI-produced academic content | 31% | Australia-wide |
| Plagiarism incidents | 19% | Reports |
National Tertiary Education Union vice-president Leah Kaufmann notes academics' low AI literacy and policy confusion: "Staff are struggling to keep up."
Learn ethical AI use in research rolesChallenges with AI Detection Tools
Tools like Turnitin flag human text (even The Bible) as AI-generated and miss paraphrased outputs. Curtin, University of Queensland, Deakin, Monash, and UTS report surges but distrust detectors. University of Sydney's Danny Liu urges: "Academics are teachers, not police—teach proper use." TEQSA recommends redesign over bans.
TEQSA AI Risks Report (PDF)Redesigning Assessments for an AI World
Experts advocate AI-resistant designs:
- Staged submissions showing evolution.
- Personalized prompts based on class discussions.
- Reflective journals on AI ethics.
- Collaborative projects emphasizing human insight.
UNSW guides AI levels per task; Sydney requires citations. By 2026, expect mandates for AI literacy modules.
Broader Implications: Degree Value and Employability
Unaddressed cheating risks devaluing degrees, with employers questioning graduates' skills. Finkel warns of reputational damage: "Degrees handed out like expensive lollies." Regional disparities worsen, as remote students face barriers. Positive note: AI integration could enhance learning if ethical.
Rate professors on AI policy handlingStakeholder Perspectives and Solutions
Universities: Invest in proctoring tech, training.
Students: Embrace transparency, cite AI.
Government: Fund redesigns via TEQSA.
Solutions: National standards, hybrid models balancing flexibility and integrity.
Future Outlook: Balancing Innovation and Integrity
By late 2026, expect widespread adoption of in-person mandates, improved detectors, and AI curricula. Partnerships like those with OpenAI could help, but vigilance is key. For academics seeking roles in this evolving landscape, explore opportunities at AcademicJobs.com higher ed jobs. Students, check career advice on thriving ethically. Share your views below—how is your uni adapting?
Photo by Martin David on Unsplash
