Australia's Fair Work Commission is facing a workload crisis caused directly by AI. From approximately 30,000 matters per year before 2023, the Commission's caseload has grown to 45,000 in 2024-25 and is projected to reach 55,000 this financial year, a 70% increase in three years driven significantly by employees using generative AI tools to draft unfair dismissal claims, general protections applications and workplace complaints.
In one case from 2026, Riley v Nuvei Australia Merchant Services [2026], an applicant submitted a claim containing fake case law citations generated by AI. The Commission published draft guidance in March 2026 on the use of generative AI in proceedings, identifying an unprecedented GenAI-driven increase in workload as the driver. It is taking submissions on whether hyperlinking obligations should extend beyond legal practitioners to all users of AI in Commission proceedings.
The AI-in-the-workplace story in Australia is not primarily a story about what employees are doing with AI. It is a story about what employers are doing with AI and what obligations already attach to those decisions, with specific and approaching deadlines that most HR and compliance functions have not fully mapped.
The Regulatory Landscape for AI in Australian Workplaces
Australia has not passed a standalone AI employment law. But the intersection of existing law with AI-assisted employment decisions creates a compliance environment that is more demanding than most employers recognise.
The NSW Work Health and Safety Amendment (Digital Work Systems) Act 2026, passed on 12 February 2026, is Australia's first AI workplace safety law. It imposes specific work health and safety duties on persons conducting a business or undertaking (PCBUs) that use AI, algorithms, automation or online platforms to allocate work. Commencement is by proclamation and the Act's full scope is still being assessed across the sector. Other states are watching. This is not a theoretical future obligation.
Fair Work Act consultation obligations apply when a major workplace change is likely to significantly affect employees. The Australian Council of Trade Unions issued a formal warning to employer peak bodies in March 2026: organisations that do not consult workers, in accordance with legal obligations, about decisions to introduce AI where it is likely to change employees' jobs or how they do them will face a coordinated ACTU response. The ACTU's position is explicit and legally grounded. Under the Fair Work Act and most modern award consultation clauses, AI introduction that significantly affects work duties, hours or job structure triggers consultation obligations before implementation, not after.
Anti-discrimination obligations apply to AI-assisted hiring, performance management and redundancy decisions in exactly the same way as they apply to human-made decisions. If an automated hiring tool filters out candidates based on protected attributes, the employer carries vicarious liability for the bias in the algorithm. The Fair Work Commission would assess an AI-driven dismissal by the same standard as a human-driven one: valid reason, fair process, human accountability.
The Privacy Act automated decision-making obligations from 10 December 2026 apply specifically to employment AI. If an AI system influences performance ratings, promotion decisions, pay decisions or redundancy selections, the organisation must disclose this in its privacy policy and ensure the decision process is documented. The right to explanation provision means that from 10 December, an employee whose promotion application is influenced by an AI system may be entitled to a plain-English explanation of how that decision was made. Organisations that cannot produce that explanation because the AI system was never adequately documented have a compliance gap.
The Four AI Employment Decision Categories That Carry the Most Risk
Not all AI use in the workplace creates equal legal exposure. Four categories of AI employment decisions carry the most immediate regulatory and legal risk for Australian employers.
AI in recruitment and hiring. Resume screening, video interview analysis, psychometric assessment and candidate ranking tools all carry discrimination risk if their outputs correlate with protected attributes. The Fair Work Commission's existing framework applies. ASIC and APRA regulated entities also face sector-specific conduct obligations around fair treatment that extend to how AI is used in hiring for roles that affect customers.
AI in performance management. Performance scoring systems, productivity monitoring, algorithmic task allocation and AI-generated performance ratings all carry Fair Work Act exposure if they influence decisions about pay, promotion, role change or disciplinary action. The NSW Digital Work Systems Act creates specific WHS obligations around algorithmically allocated work. The Privacy Act right to explanation applies from December.
AI in redundancy and restructure. Using AI to identify roles for redundancy, rank employees for retention or generate restructure recommendations creates unfair dismissal risk if the process is not demonstrably fair. An algorithm that recommends redundancy does not absolve the employer of the obligation to apply a fair process, consult with affected employees and document the valid reason for each decision.
AI-driven monitoring and surveillance. Productivity monitoring tools, communication analysis software and location tracking AI must comply with the NSW Workplace Surveillance Act requirement to notify employees of computer surveillance. Other states have equivalent obligations. The intersection of AI monitoring with privacy obligations is an area where many employers have exposure they are not aware of.
What Good AI Employment Governance Looks Like
The organisations handling this well are those that have mapped their AI employment tools against the specific legal frameworks listed above, documented the accountability structures for each tool and built consultation and disclosure processes into their AI adoption workflow.
Documenting every AI system that influences employment decisions is the foundational step. This is the AI inventory exercise that responsible AI governance requires for all AI systems: applied specifically to the employment context, it produces the evidence base that satisfies Fair Work Act, Privacy Act and WHS obligations simultaneously. Trusenta's AI Governance platform provides the use-case register infrastructure to capture employment AI systems with the specific documentation required for compliance, including the decision type, the data used, the human oversight mechanism and the accountability owner.
Establishing human oversight requirements before deployment is the second step. The advice from every major employment law firm active in this space in 2026 is consistent: maintain human involvement in significant employment decisions made using AI, particularly in hiring, firing, promotion and performance management. This is not just legal advice. It is the most practical mechanism for satisfying Fair Work Act procedural fairness requirements, Privacy Act right-to-explanation obligations and WHS oversight duties simultaneously.
Consulting workers before introduction, not after, is the third step. The ACTU has made its position clear. The legal obligation exists in the Fair Work Act and most modern awards. Organisations that treat AI tool introduction as a technology project rather than a workplace change that triggers consultation obligations are creating industrial relations exposure that is entirely avoidable.
What This Means for Your Organisation
At Trusenta, the AI employment obligations question is arriving in governance conversations with increasing frequency from HR leaders, employment counsel and Chief Risk Officers who recognise the approaching Privacy Act deadline but have not yet mapped the broader employment law landscape against their AI deployment decisions.
The practical work is achievable within the timeframe. Mapping existing AI employment tools against the four risk categories above, documenting accountability structures and human oversight requirements, and establishing a consultation process for new AI deployment in the people function can be completed well before December 2026 for most organisations. The organisations that start this quarter will have defensible documentation and operating processes in place before the obligations arrive. The ones that start in October will be completing a compliance project under pressure.
Key Takeaways
- NSW passed Australia's first AI workplace safety law on 12 February 2026, imposing specific WHS duties on PCBUs using AI to allocate work
- The Fair Work Commission's workload has grown 70% in three years driven by AI-generated claims, and the Commission published draft AI guidance in March 2026
- The ACTU issued a formal warning in March 2026 that employers not consulting workers on AI adoption in accordance with Fair Work Act obligations will face a coordinated response
- Four AI employment decision categories carry the most risk: recruitment, performance management, redundancy and workforce monitoring
- Privacy Act ADM obligations from 10 December 2026 include a right to explanation for AI-influenced employment decisions
- The governance response is an AI employment tool inventory, documented human oversight requirements, consultation process and disclosure documentation built before the obligations arrive
How Trusenta Can Help
AI Governance: Trusenta's AI Governance platform provides the use-case register infrastructure to capture AI employment tools with the documentation required for Fair Work Act, Privacy Act and WHS compliance, mapping the accountability owner, decision type, human oversight mechanism and data handling for each system.
Risk Management: Trusenta's Risk Management module supports the documentation of AI employment risks, including discrimination risk, unfair dismissal exposure, WHS obligations and Privacy Act ADM compliance gaps, with treatment plans linked to specific controls and monitoring requirements.
AI Governance Services: For organisations that need to build the AI employment governance framework before December 2026, Trusenta's AI Governance engagements deliver the inventory, accountability structures, human oversight frameworks and consultation processes that make compliance achievable before the obligations arrive.
Conclusion
AI employment obligations in Australia are not waiting for a standalone AI employment law. They are already here, embedded in the Fair Work Act, anti-discrimination legislation, state workplace surveillance laws, the new NSW Digital Work Systems Act and the approaching Privacy Act ADM obligations. The employers getting ahead of this are not waiting for legislation that gives them a checklist. They are applying the existing legal framework to their AI deployment decisions now, building documentation and processes that will satisfy obligations they can already see arriving. The employers that are not doing this will find that December 2026 arrives with compliance gaps that were entirely foreseeable from where we are sitting today.
