Australia's Fair Work Commission attributes a 70% increase in workload to AI.
TL;DR
Australia’s Fair Work Commission has initiated a process review after experiencing an estimated 70% increase in workload over three years, partly due to generative AI tools that allow more individuals to file longer, more complex, and sometimes inaccurate claims. Similar trends are observed in New Zealand’s Tenancy Tribunal and Australia’s financial complaints authority.
Australia’s Fair Work Commission has launched a review of its processes to manage what it describes as a 70% increase in workload over three years, partially attributed to the rise of generative AI tools. The commission, responsible for handling claims related to unfair dismissals, wage disputes, discrimination, bullying, and workplace sexual harassment, stated that this surge is directly impacting its capacity to provide timely dispute resolution, as noted in a statement released on Friday.
The figures illustrate the situation. Between July 2025 and April 2026, the commission received 44,039 applications with two months left in the financial year. The complete 2024–25 year recorded 44,075 applications, and the commission is projected to surpass that figure by a substantial margin.
How AI alters filing practices
The commission linked the rise in lodgments to several factors: an increase in individuals representing themselves in workplace cases, budget limitations, resource challenges, and the widespread use of generative AI tools that facilitate the creation of polished yet often generic submissions. This suggests that AI is making it easier for individuals, who might have previously deemed their cases unworthy of pursuit, to generate detailed claims quickly.
In March, the Fair Work Commission issued draft guidance stating that anyone utilizing generative AI to prepare documents for submission must disclose this fact. The guidance cautioned that AI-generated information could be incomplete, inaccurate, or fabricated. A new section for “Use of GenAI” will be added to all commission forms.
Commission's response
The commission's plan includes trialing a new approach where senior staff assist in resolving disputes informally at an earlier stage in the process, preventing cases from requiring full hearing time. They have also reassessed their application management and are contemplating the use of an AI voice agent to help manage calls to their helpline.
The irony of a tribunal inundated with AI-generated submissions seeking to implement AI tools to handle the influx is notable. Australia has already endorsed AI in other legal areas, including a government-supported chatbot that assists couples in asset division during separations. The rationale is logical: if generative AI is increasing the volume of incoming work, employing automated triage may be essential to keep up without proportionate staffing increases that budget constraints have already made impractical.
A similar issue in New Zealand
This trend is also emerging across the Tasman. Radio New Zealand recently reported that tenants in New Zealand are leveraging AI to assist with applications to the Tenancy Tribunal, resulting in additional work and delays. In one instance, a tenant utilized AI to file a claim for $40,000 related to unsafe drinking water and a broken dryer, resulting in the tribunal awarding only $80.
This case highlights a recurring challenge with AI-generated legal filings. These tools can produce seemingly authoritative arguments that may cite irrelevant legal principles, claim disproportionate damages, or reference legal frameworks from other jurisdictions. Adjudicators are then faced with the labor-intensive task of sifting through extensive material to determine what is pertinent.
Financial complaints parallel
The Australian Financial Complaints Authority (AFCA), which manages disputes in financial services, reported increased usage of AI in how consumers engage with financial firms and lodge complaints. A spokesperson acknowledged that while AI can assist some individuals in articulating their concerns, it can also yield complaints that include irrelevant or inaccurate information or employ legal arguments not applicable under Australian law.
AFCA encourages individuals to keep their complaints straightforward, as lengthy AI-generated submissions complicate and delay the resolution process, forcing staff to sift through excessive material to identify core issues. This suggests an acknowledgment that verbosity does not equate to a stronger case, and AI's propensity for generating lengthy, confident-sounding content can hinder dispute resolution.
Access to justice or just noise?
The central tension of the issue is noteworthy. AI tools can enhance access to legal processes for those unable to afford legal representation, a benefit promoted by governments and AI companies alike. However, when these same tools lead to submissions that are lengthier, less accurate, and more complicated to navigate than those produced by humans without assistance, the overall effect may be a slowdown in the system for all users.
Australia’s Fair Work Commission is the first significant tribunal to publicly recognize generative AI as a factor in its workload challenges, a trend likely to be mirrored by other institutions. Any entity that accepts written submissions from the public is now grappling with these dynamics; AI makes text generation easy, while establishments designed for human-pace filings struggle to manage the results.
Other articles
Australia's Fair Work Commission attributes a 70% increase in workload to AI.
Australia's Fair Work Commission indicates that generative AI tools have contributed to a 70% rise in workload over three years. New Zealand and financial complaints organizations are observing a similar trend.
