AI now engages established workplace laws
There is no single Australian statute regulating AI in employment. Instead, an employer’s use of AI may be governed by various legislation or regulations, including relating to work health and safety, privacy, discrimination and workplace surveillance. The applicable obligations depend on the function performed by the system and the effect of its use, rather than whether the product is described as ‘AI’. This sits within the broader regulatory landscape for AI in Australia, which we explored in our recent insight, Prime Minister announces national AI regulation: Key takeaways for investors, developers and operators.
Employers should focus on systems that influence employment outcomes. Candidate screening tools may create privacy and discrimination risks. Automated rostering, workload allocation and productivity monitoring may affect consultation obligations and impact psychosocial safety. Systems used in performance management or disciplinary processes may create additional legal risk if inaccurate, incomplete or discriminatory outputs are adopted without proper human review. Depending on the circumstances, this could affect whether there was a valid reason for disciplinary action or dismissal, whether the employee was afforded procedural fairness and whether the decision involved unlawful discrimination or adverse action.
NSW legislates new WHS requirements for digital work systems
The Work Health and Safety Amendment (Digital Work Systems) Act 2026 (NSW) received assent on 18 February 2026. It will commence on a date yet to be announced by the NSW Government. Once the substantive amendments commence, they will introduce express provisions concerning ‘digital work systems’, including systems that use algorithms, artificial intelligence (AI), automation and online platforms. The amendments will impose specific requirements concerning health and safety risks arising from the use of digital work systems and the allocation of work by those systems. They will also expand WHS entry permit holders’ rights in relation to relevant systems.
Existing WHS duties already require a person conducting a business or undertaking to eliminate risks arising from workplace technology. If elimination is not reasonably practicable, they must minimise those risks so far as is reasonably practicable. Safe Work Australia identifies potential risks associated with AI and digital technologies to include work intensification, fatigue, reduced job control, poor organisational justice, inadequate training and intrusive monitoring.
Before introducing or materially changing a digital work system, employers should assess and document the associated risks, consult affected workers and any health and safety representatives as required and implement appropriate controls. The assessment should address the system’s purpose and the data and assumptions on which it relies on. It should also address the system’s effect on workload, pace, autonomy and monitoring and whether its outputs can be challenged or overridden by an appropriately authorised person.
The Fair Work Commission imposes disclosure and verification requirements
The Fair Work Commission’s Guidance Note on the use of generative AI took effect on 20 October 2026 and applies in all Commission cases. A person who uses generative AI to create, write, edit or otherwise prepare a document must disclose that use and explain how the technology was used. They must personally check the information is accurate, relevant and supported by real evidence and legal authorities and state in the document that those checks have been completed.
For a witness statement or declaration, the witness or declarant must also ensure the document is based on their own knowledge, reflects their own words and is true to the best of their knowledge. Legal practitioners and other professional representatives must include hyperlinks to any case law referred to in a document prepared using generative AI.
A failure to comply may result in the Fair Work Commission giving the document less weight or disregarding it, making a costs order or dismissing the application or case.
Similar practical issues may arise when employers receive lengthy AI assisted grievances or demands. The use of AI does not determine whether the underlying complaint has substance. Employers should isolate the factual allegations, seek particulars where necessary, preserve relevant evidence, assess whether any immediate risk controls are required and provide a proportionate response to the matters raised.
New privacy transparency obligations apply to automated decisions
From 10 December 2026, new transparency obligations under the Privacy Act 1988 (Cth) will require Australian Privacy Principle (APP) entities to disclose in their privacy policies certain information about the use of computer programs in decisions that could significantly affect individuals. We examine the scope of these automated decision-making (ADM) obligations in detail in our recent insight, As breaches grow and frontier AI sharpens cyber threats, transparency on ADM has never mattered more.
For employers, these changes are particularly relevant to systems used to screen applicants, rank employee performance, recommend remuneration or promotion outcomes or otherwise materially influence decisions affecting individuals. The application of the obligations to existing employee records will require consideration of the employee records exemption and the way in which the relevant information is collected, held and used.
Priorities for employers in 2027
Employers should maintain a current register of AI and digital work systems, including functions embedded in HR, payroll, recruitment, productivity and monitoring products. Each system should be classified by reference to its purpose, the information it processes and its potential effect on workers and other individuals.
Systems that materially influence employment decisions should be subject to clearly documented ownership, appropriate privacy and WHS assessments, consultation where legally required, testing for errors and discriminatory outcomes, secure data handling, meaningful human review, a process for challenging disputed outputs and periodic review.
Human review must be substantive rather than merely formal. The person responsible for the final decision should understand the basis of the system’s output, test it against reliable evidence and document the independent reasons for the decision.
Let's chat: Work
Published
19 November 2026This update does not constitute legal advice and should not be relied upon as such. It is intended only to provide a summary and general overview on matters of interest and it is not intended to be comprehensive. You should seek legal or other professional advice before acting or relying on any of the content.