The UQ underpayments reached about A$11 million, and they ran between 2017 and 2026. In September 2026, the University of Queensland signed an Enforceable Undertaking with the Fair Work Ombudsman after underpaying 16,382 current and former staff. The university reported the problem itself.
Nothing in the public record says this was deliberate. That’s what makes it worth reading. It shows how a small payroll error can grow into a large bill without anyone meaning it to. Here’s how it grows.
It starts with a setting
According to the Fair Work Ombudsman, the underpayments came from incorrect payroll system settings, clerical errors and other oversights (HCAMag, September 2026). None of those is dramatic. Each is the kind of thing a busy team can miss.
It repeats in every pay run
A wrong setting doesn’t make one mistake. It can make the same mistake every time it runs. At UQ, the underpayments ran between 2017 and 2026 (Fair Work Ombudsman, 1 September 2026).
It adds up across thousands of people
The total reached 16,382 current and former staff. The university underpaid about A$9.1 million in wages, about A$427,000 in interest, A$1.02 million in superannuation and about A$315,000 in interest on that superannuation. Among the first 13,514 staff repaid, individual payments ranged from less than A$1 to A$41,643 (HR Leader, September 2026).
It stays hidden until someone looks
UQ found the underpayments through an external review of its wage compliance, then reported them to the regulator. The Fair Work Ombudsman said it acknowledged that UQ reported the underpayments and has worked to fix them and prevent more.
In April 2025, UQ also told the regulator about a review of casual staff pay covering part of 2024 (enforceable undertaking).
It ends with a promise and a bill
The Enforceable Undertaking, signed on 1 September 2026, requires payments to the remaining staff. UQ also committed to tell the regulator about its improved systems, train relevant staff, run an independent compliance audit, consult staff and the National Tertiary Education Union on complaints systems, and strengthen oversight through its Senate Risk and Audit Committee. It was the 15th such undertaking with a university.
Campus Review reported that UQ had already repaid more than A$9.6 million. Two days later, Murdoch University signed its own undertaking, covering A$4.9 million for more than 2,100 staff.
The statistic about everyone else
In May 2026, ADP, a payroll provider, released a poll of 1,019 Australians run by the research agency Antenna. 31% said they had been underpaid in their usual pay period at least once, losing an average of A$1,131 each time. 25% said they had noticed at least one mistake on their payslip in the past 12 months.
That’s a lot of people who can see a problem. The question is whether the people who can fix it ever hear about it.
UQ’s position
The Fair Work Ombudsman acknowledged that UQ reported the underpayments and has worked to fix them and prevent future problems. We could not find a public statement from UQ to quote.
What MoodyBot could have saved in the UQ underpayments
MoodyBot, the AI layer inside Moody At Work, reads anonymous daily mood check-ins across teams. It doesn’t run payroll or check pay rates. Instead, it tracks how people feel at work every day, with no names attached.
In a university, MoodyBot could have flagged staff feeling undervalued or unsure about their pay. It could also have shown check-in reasons clustering around payslip confusion, casual staff feeling overlooked or not knowing who to tell. Finally, it could have shown which areas felt most uncertain.
A payroll error is hard to see from the top. A steady pattern of “my pay looks wrong” in anonymous data is much easier to notice.
| Risk | Publicly reported loss | Model assumptions | What MoodyBot could have saved |
|---|---|---|---|
| Interest on underpayments | About A$742,000 (A$427,000 on wages plus A$315,000 on superannuation) | 50% to 80% avoidable with earlier detection (interest grows with every year an error runs); 30% to 50% chance early signals lead to action | A$111,300 to A$296,800 |
| Wages and superannuation owed | About A$10.1m (A$9.1m wages plus A$1.02m superannuation) | Money owed to staff either way | Not counted |
| Audit and system fixes | Independent compliance audit and improved systems required | Cost not disclosed | Not quantified |
| Staff trust | 16,382 current and former staff underpaid | Not measurable from public data | Not quantified |
| Total quantified | A$111,300 to A$296,800 |
How we built the numbers: A$427,000 plus A$315,000 gives A$742,000 in interest. Then A$742,000 x 50% x 30% gives A$111,300, and A$742,000 x 80% x 50% gives A$296,800. Wages and superannuation are owed to staff in any case, so we don’t count them as a saving. The real cost is the years that thousands of people were short.
What every leader can learn
Pay errors rarely come from bad intent. They come from systems nobody checks often enough. Therefore, ask more than “is payroll right?” Ask “how would we find out if it weren’t?”
Pay problems also built up quietly at Coles, as we wrote in the Coles underpayment story.
UQ underpayments FAQs
How much did UQ underpay staff? About A$11 million, including interest and superannuation, to 16,382 current and former staff between 2017 and 2026, according to the Fair Work Ombudsman.
Did UQ report the problem? Yes. UQ reported the underpayments, and the Fair Work Ombudsman acknowledged that when it accepted the Enforceable Undertaking on 1 September 2026.
How common are pay mistakes? According to an ADP poll in May 2026, 31% of Australians say they have been underpaid at least once.
Sources: ADP and Antenna poll (14 May 2026); Fair Work Ombudsman (1 September 2026); HCAMag (September 2026); HR Leader (September 2026); Campus Review (September 2026).
All factual claims are sourced from publicly available reporting. The Fair Work Ombudsman acknowledged that the University of Queensland reported the underpayments and has worked to fix them, and we could not find a public statement from the university. MoodyBot projections represent an independent illustrative probability model based on publicly reported information and are not guaranteed outcomes.