The university AI fight has moved beyond policy papers. About 2,000 University of Sydney staff walked off the job on Wednesday, raising concerns about job security, workload, educational quality, and who gets to decide how artificial intelligence enters the classroom.
From 7am, about 500 people were stationed at picket lines at each entrance to the university’s inner Sydney campus. Staff summed up the mood with a blunt message: “We have had enough.” The University of Sydney told students that “almost all” scheduled classes would move online because of the strike.
The dispute centres on demands from the university’s National Tertiary Education Union branch, the largest in Australia. The union is asking for best practice in AI use, due diligence, and protections around systems that could alter teaching work without adequate safeguards.
“We are asking for best practice, we are asking for due diligence. Not their own commercialisation aims and certainly not the interest of venture capital,” said Vince Caughley, the New South Wales division secretary of the NTEU. He described AI as the most “fundamental transformation” education had seen, while the union said it was confident of reaching an agreement with management within the year.
A trust problem before an AI problem
The strike lands on top of a severe trust gap inside the University of Sydney. In its faculty of arts and social sciences, 0% of staff agreed with the statement: “I trust the University of Sydney to act in the best interests of staff, students and the broader community.”
That figure gives the argument its real shape. This is not only a quarrel over software; it is a fight over whether staff believe university leaders will protect teaching standards and working conditions when commercial interests enter the room.
Peter Chen, president of the University of Sydney NTEU branch, said: “Higher education is the canary in the coalmine. Where we go on AI … will travel through the rest of the economy.” The warning is aimed beyond the campus, because universities are testing work arrangements that may later appear in other industries with fewer institutional brakes.
The university said it was “continuing to engage” with the union and maintained that industrial action was “premature”. It supports many of the unions’ claims but wants them implemented through policies rather than the Enterprise Agreement, creating a basic disagreement over whether AI protections should be binding employment terms or management rules that can change outside bargaining.
Macquarie’s chatbot puts the argument in front of students
Macquarie University has made the debate tangible through Virtual Peer, an AI chatbot used to help teach two compulsory psychology subjects. The system interacts with students during coursework, asks questions, and guides exercises; the university says it is not a replacement for conventional teaching and that its activities are optional.
Students still complete assessments through research assignments and examinations. Macquarie plans to bring classroom teaching back in 2027, alternating between in-person and online formats, but students have already expressed concern that online courses replaced face-to-face teaching and left them short-changed.
The change replaced optional weekly Zoom tutorials led by paid academics with Virtual Peer, alongside digital materials and videos. One student described the experience this way: “Instead of a group discussion or a tutor teaching you, it’s a class activity done with a robot. I feel like I’m teaching this robot how to take jobs away from my teachers.”
Macquarie says educators developed the chatbot alongside the university and that it is intended to complement their work rather than eliminate it. Dr Alison Barnes is an academic at Macquarie University, while the institution has stated that students should also work through online readings, quizzes, and videos.
The University of Sydney has expanded its own AI use through Cogniti, an education-focused generative AI platform used by hundreds of academics for chatbots and scenario-based learning. Cogniti is available to institutions in Australia, New Zealand, and the Netherlands, turning campus experimentation into part of a broader education market.
For staff, the concern is that AI rollouts could worsen existing workload problems while universities describe the tools as support. For students, the question is more direct: whether an optional chatbot adds useful practice or quietly replaces the paid academic who once answered questions, led discussion, and made the course feel like education rather than content delivery.
Prof Laura Shepherd, a member of the University of Sydney’s bargaining team, said students “come here … for quality and access to research leaders.” That expectation is now colliding with institutions that want AI in the teaching model before staff have secured clear rules for its use. Universities may call that innovation. The people standing at the gates call it a job-security dispute.
Based on
- ‘We have had enough’: thousands of University of Sydney staff walk off the job over AI and job security — theguardian.com
- AI chatbot helps teach online-only psychology classes at Macquarie University | AI (artificial intelligence) | The Guardian — theguardian.com
- Australian university replaces in-person psychology classes with AI chatbot for two subjects | The Independent — independent.co.uk
