Rise of AI “slop” Fair Work claims: new hope for employers?

Have you noticed an increase in employment claims by current or former employees in the Fair Work Commission (Commission)?

Have you noticed strikingly similar language, bolding, unnecessary hyphenation and formatting commonly associated with AI‑generated documents?

At first glance, these claims appear well‑structured and coherent, giving the impression of a clear and compelling case of adverse action or unfair dismissal. On closer review form does not reveal substance. What appears to be a detailed argument is, in many instances, superficial padding obscuring an otherwise weak claim with little prospects of success.

You’re not the only one who has noticed. His Honour, Justice Hatcher, President of the Fair Work Commission, recently gave a presentation to the Victorian Bar Association on the rise of AI in Fair Work Commission proceedings.

The AI burden on employers and the system

According to his Honour, lodgements in the Commission have risen from around 30,000 matters annually before 2023 to a projected 50,000–55,000 in 2025–26, with

• Applications for general protections (GP) involving dismissal increasing 62%;

• Applications for GP not involving dismissal up 135%; and

• Unfair dismissal applications up 41% compared to the three-year average.

This has increased the Commission’s total workload by over 70% in the space of three years. This has been keenly felt by employers and their representatives alike, with significant delays in processing matters within previous expected timeframes.

Justice Hatcher reasonably deduced this significant workload increase was due to the accessibility of AI tools to litigants, particularly self-represented litigants, concluding that:

“The clue that the growth in lodgments was AI-driven first became apparent by the widespread use of AI-generated language in the applications being filed. Once you learn what this looks like, it becomes pretty easy to spot.”

Justice Hatcher put AI to the test by himself and the results were alarming:

“I told ChatGPT that I had been dismissed and asked it what I could do. After I provided it with few basic facts, including the reason for my dismissal and my view that I had been dismissed because I made a complaint a few years earlier, ChatGPT prepared an application under section 365 of the FW Act that was in a form ready to file together with a witness statement which contained a substantially-invented story about my dismissal. It also told me that a ‘realistic scenario’ was that I would obtain in the range of $15,000–$40,000 compensation. This took less than 10 minutes. And this was for a putative case which, on the facts I provided to ChatGPT, could not be assessed as having reasonable prospects of success.”

Not only has it become more burdensome on the employer to defend claims that are lacking in substance and have no reasonable prospects of success, but also “it has become harder to quickly and efficiently identify and dispose of cases which do not have reasonable prospects of success” due to the use of AI dressing up baseless claims in a legalistic manner.

This results in more time and resources spent by employers and the Commission alike deciphering whether the claims have any merit in the first place.

What the Commission proposes to do

The Commission has indicated that it will be looking into reforming its procedures “directedly targeted at AI use”. That may soon include:

• A requirement to disclose AI use in the required AI disclosure in the preparation of applications and witness statements;

• Acknowledge that applicants have checked the correctness of details, including facts and case law in the document, while legal practitioners and paid agents will also be expected to provide hyperlinks for case law they cite; and

• Ensuring that witnesses or declarants to witness statements or declarations certify that they have checked the documents and ensured they are based on their own knowledge and are true to the best of their knowledge.

With respect to GP applications, the Commission has already completed its first stage in implementing procedures that seek to address some of these issues.

This is long-overdue assistance in a jurisdiction which has too long presented a loophole for applicants represented by paid agents to baselessly seek eye-watering sums in compensation (including for “hurt, distress and humiliation” unsupported by any medical evidence), relying on the reverse onus of proof placed on the employer to disprove these claims. Clearly, this has only been made worse by AI, as demonstrated by Justice Hatcher’s own test of ChatGPT.

The main features of these reforms are:

• New general protections application and response forms which will require a more rigorous articulation of the applicant’s case that the general protections provisions have been contravened;

• An increase in the Commission’s capacity to decide and dismiss extension of time applications which do not have reasonable prospects of success on the papers without requiring the employer to respond;

• Clear notices to parties that conciliation conferences are not a substitute for a hearing and are only for the purpose of discussing settlement;

• A more rigorous approach to allowing representation at general protections conciliation conferences and shorter conferences concerned only with possible settlements; and

• If one or both parties make it clear that they do not wish to enter into a settlement, that the conference will be terminated immediately.  

What does this mean for employers?

The internal, process-based reforms implemented and proposed by the Commission will hopefully begin to tip the scales back to a fair system for employees and employers alike.

However, in the face of the onslaught of AI-assisted and meritless claims, this may take some time, during which employers will still face an increased burden in defending these claims.

Some strategies employers should consider in this environment are:

• Taking a more critical view of applications filed in the Commission, particularly GP applications where the applicant is self-represented; and

• Being prepared for short, targeted conciliations focused on settlement discussions rather than the merits of a claim.

We are here to assist navigating this mad new world, including to provide a critical review of any applications received, and to assist with strategy in defending claims.

Authors: Hannah Lawson & Giuliana Quibell

Contributing partner: Darren Gardner

This publication is intended as a source of information only. No reader should act on any matter without first obtaining professional advice.