Armin Alimardani, a senior lecturer in law and technology at Western Sydney University, argues that replacing human academics with algorithms risks creating a destructive slop-cycle where AI-generated student work is assessed by automated systems.
Several institutions across Australia currently permit academic staff to use artificial intelligence to assist with marking assessments and providing student feedback. Western Sydney University, the University of Newcastle, Deakin University, RMIT in Victoria, and the University of Adelaide all provide a qualified green light for staff to incorporate AI tools into their grading workflows.
At the University of Newcastle, students retain the right to opt out of AI-assisted assessment. Deakin University permits staff to use the technology to improve efficiency and support administrative activities, though institutional guidelines explicitly prohibit AI from assigning final grades. A Western Sydney University spokesperson stated that the institution maintains a human-centred approach where the awarding of marks and feedback remains strictly the responsibility of academic staff.
Conversely, several high-ranking institutions draw a firm line against automated grading. The University of New South Wales, the University of Melbourne, and the University of Sydney prohibit generative artificial intelligence in the marking process entirely, despite allowing its use in other classroom activities. Other institutions, including Monash University, the University of Queensland, and the Queensland University of Technology, are currently reviewing their assessment frameworks or rolling out broader transformation projects.
The Risk of Verification Drift and Academic Redundancy
Alimardani warns that relying on AI to speed up assessments introduces severe oversight vulnerabilities. He describes a phenomenon he calls verification drift, where overworked academics grow complacent after reviewing a string of accurate AI-generated assessments and eventually miss hallucinations or errors slipping through as facts. Such scandals would quickly erode institutional reputations and damage student trust.
Staff unions and academics express deep apprehension about job security, particularly for sessional academics who rely on marking as a primary source of income. A Queensland University of Technology academic, speaking on the condition of anonymity, noted that a recent university communication regarding an assessment uplift project sounded like corporate speak signaling a reduction in human marking. Sessional staff worry their roles face redundancy as universities pursue administrative efficiency.
Student Backlash Against Automated Assessment
Students push back against the prospect of algorithmic grading, arguing it undermines the core value of higher education. Alex Cameron, a creative and professional writing student and editor at Glass Magazine, states that paying thousands of dollars for a degree marked by software defeats the purpose of attending university. Cameron insists that students expect human engagement, authentic writing, and personalized feedback from qualified academics rather than automated evaluations.
Australian universities permit staff to use artificial intelligence for grading
Western Sydney University, the University of Newcastle, Deakin University, RMIT, and the University of Adelaide permit staff to use AI tools for specific marking and feedback tasks under strict institutional guidelines.
Do universities allow students to use AI for assignments?
Policies vary widely, but universities maintain strict academic integrity standards regarding student AI use while warning against the hypocrisy of using automated tools to grade work that students are penalized for generating.
What is verification drift in AI grading?
Verification drift occurs when human supervisors grow less vigilant over time while overseeing automated systems, increasing the likelihood that AI errors and hallucinations go unnoticed during the assessment process.
Keep reading