'We Have Had Enough': Thousands of University of Sydney Staff Walk Out Over AI and Job Security

“We have had enough.” With those words, thousands of staff at the University of Sydney walked off the job on Wednesday, September 2, mounting one of the most significant industrial actions the Australian university sector has seen in years — and the first to put artificial intelligence governance at the center of a live workplace dispute.

Roughly 2,000 academics and professional staff represented by the National Tertiary Education Union (NTEU) staged a 24-hour strike after nearly five months of stalled Enterprise Agreement negotiations. Hundreds manned picket lines across the Camperdown and Darlington campuses from dawn. The immediate effect was visible in the timetable: almost all scheduled classes moved online, while others were rescheduled or cancelled outright. Professional staff represented by CPSU NSW did not join the walkout but were instructed not to cover the duties of striking colleagues.

What the strike is actually about

On the surface, the dispute looks like a familiar bundle of enterprise-bargaining demands: a 20 per cent pay rise over three years, 17 per cent superannuation for casual staff, workload caps, and stronger job security through decasualisation. But the strike’s distinctive fault line — and the reason it has drawn international attention — is the union’s demand that legally binding AI protections be written directly into the Enterprise Agreement, the negotiated employment contract that governs working conditions.

The NTEU wants provisions that would protect staff against adverse effects of AI systems in the workplace. In practice, that means requiring consultation before significant AI-driven workplace change, safeguards against automated decision-making in employment matters, and constraints on how AI can be deployed in teaching, marking, and assessment. As NTEU official Vince Caughley pointed out, AI chatbots are already helping teach online-only classes at other universities — a fact that makes the abstract debate about “AI governance” suddenly concrete for the people whose jobs it touches.

The university’s position is subtler than a simple refusal. Management says it supports many of the same objectives, including what it calls a human-centred approach to ethical AI, and notes that its AI principles and guardrails were developed through community consultation. Its argument is about where those rules should live: in university policy, which management controls and can revise, or in the Enterprise Agreement, which is legally binding and can only be changed through negotiation. According to the NTEU’s bargaining updates, management rejected the union’s proposed protections at the bargaining table as far back as July, offering instead “policy-based governance and limited consultation” — a formulation staff read as unenforceable comfort.

Why this matters beyond Sydney

The disagreement, at its core, is about enforceability and control: who gets to set the rules for AI at work, and who can change them later. That question is arriving at every workplace in the economy, but universities are hitting it first and hardest, for three reasons.

First, universities are simultaneously the developers, deployers, and victims of AI. The same institutions building AI research programs are also experimenting with AI tutors, automated marking, and chatbot-led teaching — while employing the workforce most likely to be displaced or reshaped by those tools. Sydney’s staff are not opposing AI in principle; they are opposing a world where management can deploy it unilaterally while the employment contract stays silent.

Second, Australia’s enterprise-bargaining system gives this fight legal teeth. Provisions written into an Enterprise Agreement carry the force of law under the Fair Work Act, breaching them can trigger disputes before the Fair Work Commission. Policy documents carry no such weight. The union’s calculation is simple: an AI policy can be rewritten by a future vice-chancellor; an EA clause cannot. Management’s counter-calculation is equally simple: hard-coding technology rules into a multi-year contract risks freezing the university into outdated constraints as the technology evolves.

Third, the strike is a preview of bargaining fights brewing across knowledge industries. If AI-driven workplace change becomes a mandatory consultation trigger in one of Australia’s most prestigious universities, it sets a precedent every other employer will have to reckon with. The NTEU has already flagged worker-centred AI regulation as a national campaign priority, and union members at other institutions — including Western Sydney University, where staff recently won consultation rights over AI — are watching closely.

The numbers behind the dispute

The strike did not materialise out of nowhere. The current bargaining round opened in April 2026, and the union’s log of claims — a flat 20 per cent salary increase over three years, 17 per cent superannuation contributions for casual staff, decasualisation targets, and enforceable AI provisions — has been the subject of increasingly fraught negotiations since. Prior industrial actions, including an earlier strike on August 17, failed to break the impasse. Union members voted overwhelmingly to endorse protected industrial action, and Wednesday’s 24-hour stoppage was the result.

The University of Sydney issued a formal response to the industrial action on September 2, reiterating that negotiations remain ongoing and that it “remains committed to reaching a new Enterprise Agreement.” The statement emphasised the university’s existing AI governance frameworks and its willingness to continue consultations — but stopped short of conceding the central demand that AI safeguards become contractually enforceable.

For students, the timing is painful: the strike lands in the middle of the teaching semester, with classes moved online, rescheduled, or cancelled. For the sector, the stakes are larger. After nearly five months of bargaining, the key question is whether shared rhetorical support for AI safeguards can overcome the fundamental divide over who writes the rules, who enforces them, and where they live.

The answer will reverberate well beyond one campus. As one picketing academic told the Guardian, staff are not asking to stop the future — they are asking for a seat at the table where it is written. Whether that seat comes with contractual force or remains a policy promise is now the central question of the AI workplace debate, and Sydney has become its first real battleground.