AI tools are supercharging employee claims, exposing workplace compliance gaps, and forcing employers to rethink how they manage legal risk
AI has redrawn the battlelines of Australian employment law. Workers across the country now have access to powerful tools capable of stress-testing payslips, scrutinising grievance processes, and filing tribunal claims at a speed that has left employers scrambling – and the nation's workplace umpire overwhelmed.
The Fair Work Commission (FWC) is on track to record a 70% increase in total workload over just three years by the end of the 2025–26 financial year, with generative AI tools identified as a principal driver of that surge.
From roughly 30,000 matters per year prior to 2023, the FWC's caseload jumped to 45,000 in 2024–25 and is expected to exceed 50,000 in the current financial year. Unfair dismissal claims have grown 41% between 2022–23 and 2024–25, with even greater growth recorded across general protections claims.
For Brittany Byrne, partner and solicitor at Citation Legal, the shift has been both sudden and profound.
"We thought Covid was groundbreaking and it was in our space," Byrne said. "I think we're the only area of law that really thrived during those two years post-Covid and through Covid. And now we are somehow in a very different but equally as impactful period of AI. The reality is that it's honestly hit within the last two to three months."
Speed is the biggest problem
The practical effect of AI on employment claims is not just volume – it is velocity. Where employers and their legal teams once had days or weeks to formulate a response, they are now confronted with employee-generated correspondence within hours of an incident.
"What we're seeing is responses within a matter of minutes or hours," Byrne said. "What that does is it's hitting the tennis ball back over the net and our client has to react and deal with it promptly where it used to have weeks or days in between to consider its position."
The acceleration is also reshaping the profile of who brings claims. The number of self-represented litigants with unfair dismissal claims increased rapidly from early 2025, with those applicants coming to the commission with little or no workplace relations experience and often "informed by content generated by AI," according to FWC general manager Murray Furlong.
Byrne confirmed that self-represented applicants now represent the overwhelming majority of employee-related claims she encounters in both the commission and the courts.
"That is costly to businesses," she said. "And as lawyers we have to adapt to ensuring that what we're doing to manage self-represented applicants is fit for purpose."
A shift in power dynamics
The surge in claims is not simply a procedural headache – it reflects a fundamental shift in workplace power, particularly for small and medium enterprises.
Karlie Cremin, CEO of Dynamic Leadership Programs Australia (DLPA), works with employers navigating the early stages of disputes and said the psychological impact on smaller businesses has been significant.
"I think it's a highly stressful thing for a lot of employers, particularly SMEs," Cremin said.
"That's not to say it's not a stress for large employers too, but I think SMEs have less levers to pull in terms of responding to it. And as a lot of people are saying, part of the stress is that it is novel, it is emerging and we're not quite sure where the endpoint is going to be."
Cremin pointed to a structural imbalance that AI has magnified: "It's always kind of been the challenge in that area – it's a lot more expensive for an employer to defend those claims than it is for an employee to bring them," she said. "AI has accelerated that."
Both Byrne and Cremin flagged a further risk that is frequently overlooked: the removal of human judgement from what are essentially human conflicts. Cremin noted that AI-assisted claims, fired back and forth between parties, eliminate the moments of de-escalation that have historically resolved most disputes before they reach a hearing.
"The real danger of it is that both from claim side and respondent side, it takes out the human element, which is usually the point of de-escalation," she said.
"A small misunderstanding – because now we're just in this web of legal notices going back and forth – we've missed the opportunity just to have a chat."
What employers are getting wrong
According to Citation Group’s Workforce Pulse 2026 report, just 29% of businesses using AI strongly agree it is being used safely and beneficially, while 42% have identified payroll errors despite 87% believing their payroll is accurate.
On top of this, 44% have no clear HR strategy or formal people processes in place at all.
Byrne said payroll compliance is the single greatest area of legal risk for employers in the current environment.
"An employee can upload a copy of the modern award, a copy of their pay slips, a copy of their timesheet to a large language model, and two or three prompts later there is a spreadsheet that sets out what the entitlements were and where the employer had fallen short," she said.
"They then tell their colleague, who tells another colleague, who tells the union, and that problem grows exponentially."
The same dynamic applies to grievance procedures. Employees are now able to upload a company's own grievance policy alongside their correspondence and ask AI to identify where the employer has departed from its stated process. "The gaps are extensive," Byrne said.
Byrne also raised a critical and widely misunderstood risk around legal privilege. Employers tempted to substitute AI tools for legal counsel are inadvertently creating discoverable records.
"Once it's out there, you've created a record that could go to the other side," she said. "When you have a lawyer involved, of course you're creating legal privilege and so you can tell your lawyer anything. You can't do that with AI."
The court and commission response
The FWC has not been passive. Commission President Justice Hatcher introduced a proposed Guidance Note on the use of generative AI in commission cases as part of a package of reforms aimed at managing an "unprecedented GenAI-driven increase" in the tribunal's workload.
Proposals include requiring applicants to disclose whether AI has been used and to hyperlink case citations so authorities can be verified, with non-compliance potentially resulting in dismissal of an application or an order for costs.
The Federal Court has also issued guidance notes imposing accountability obligations on all parties, not just practitioners, in relation to AI use in proceedings. Byrne said many employers remain unaware of how strict those parameters are.
"Whilst it might be tempting for employers to cut out the lawyer, they might think, well, the applicant's self-represented, so we can do this too, by AI," she said.
"There's a whole lot of hoops and scrutiny from the court by way of their practice note and the rules they're putting in place to ensure that you are identifying where you're using AI, you are reporting that to the court and you're accountable to the court."
What HR leaders should do now
Both Byrne and Cremin offered practical guidance for HR leaders facing this new reality.
Byrne identified three immediate priorities. The first is a comprehensive audit of documentation and internal systems, because AI will surface deficiencies in performance management records, contracts, and payroll that may previously have gone undetected. The second is payroll compliance – ensuring that every employee's entitlements are correctly calculated against applicable modern awards. The third is manager training.
Cremin framed the challenge in broader terms. Artificial intelligence knowledge has become a core professional skill for HR practitioners in a way that was not anticipated five years ago.
"The first thing is to really put the energy into educating yourself and consciously building that skill," she said.
"The other thing is to really focus on the value of human interaction. The more human a workplace – the more connected and the more people associate you as a three-dimensional person with your own ideas and feelings – the less likely they are to bring at least a vexatious type of claim."
The consensus from those at the coalface is clear: the window to act before AI-assisted claims arrive at the door is narrowing.
The employers who audit their systems, train their people, and build workplaces grounded in human connection will be best placed to weather the future of employment law.
