By The Fair Work Commission has issued new rules requiring anyone involved in a Commission case to disclose when generative AI was used to prepare documents and to have a person verify the facts. The requirements take effect on 20 October 2026 and apply to employers, employees, HR teams, self-represented workers, legal practitioners and professional representatives, according to HRD America.
The rules sit in a new Guidance Note titled “Use of generative artificial intelligence in Commission cases” and were released with a President’s Statement from Fair Work Commission President Justice Adam Hatcher. The Commission’s workload has grown by more than 70 per cent in three years, a rise Justice Hatcher linked to the spread of GenAI tools rather than labour market conditions.
What the rules require in practice
The core obligation is twofold. First, anyone using generative AI to prepare an application, submission, witness statement or other document lodged in a Commission case must state that GenAI was used and how. Second, the document must be checked so that every fact and piece of evidence exists, every case, statute and article exists and supports the position claimed, and every quote comes from its stated source. The party must then declare that the check has been done.
The checking must be done by a person. Asking the same or a different GenAI tool to review the document does not count. This is the point that will most likely trip up self-represented workers and small employers who assume a second AI pass is sufficient.
Using GenAI only to check spelling or grammar, or to format a document, falls outside the rules. That is a narrow but useful carve-out for routine drafting tasks.
Witness statements and declarations carry a stricter standard. They must reflect the witness’s own knowledge and own words, and be declared true to the best of their knowledge. A statement drafted by AI from a witness’s dot points may not satisfy this if the phrasing is not the witness’s own.
Legal practitioners and professional representatives must also hyperlink every case they cite. The Guidance Note’s definition of a professional representative includes a human resources adviser employed by the party they represent, which means in-house HR staff appearing in Commission matters are caught by the same obligations as external lawyers.
Consequences and the cost risk for employers
Non-compliance can lead to documents being given less weight or disregarded, a party being ordered to pay the other side’s costs, or a case being dismissed. For employers, that cuts both ways. An unchecked AI-drafted response can weaken a defence, but an employer can also point to an applicant’s non-compliance when arguing for costs.
Tinashe Makamure, special counsel at Barry Nilsson, told HRD America the Guidance Note shows tribunals are “embracing it as a new reality of litigation” while also regulating it to “stem the flow of voluminous misconceived material being filed without consequence.” He added that the increase in time is “causing delays across the board and is disadvantaging those who do have legitimate claims.”
Brittany Byrne, practice leader at Citation Legal, said the rules “cut both ways” for employers. “Before anything goes to the Commission, a person, not another AI tool, needs to check every fact and legal reference,” she said.
The Guidance Note also warns against entering personal or confidential case information into public GenAI tools such as ChatGPT, Claude, Copilot or Gemini. For HR teams, that means employee records, payroll data and disciplinary material should never be pasted into a public tool to draft a response.
What HR should do before October 2026
The first step is to map where AI is already used in the HR function. Citation’s Workforce Pulse 2026 survey found 73 per cent of large businesses and 65 per cent of medium businesses use AI at work, yet only 29 per cent of those users strongly agree it is being used safely and beneficially. That gap suggests many HR teams have AI in place without clear guardrails.
Byrne’s advice is to keep original records and make sure witness statements are in the witness’s own words. HR should also never put employee details or confidential information into public GenAI tools.
A practical compliance workflow for any document heading to the Commission would look like this: identify whether GenAI was used and which tool; record what it was used for; verify every factual assertion against primary sources; confirm witness statements reflect the witness’s own words; complete the disclosure statement; and retain records of the verification.
The rules do not say disclosure alone will reduce the weight given to a document. The Commission has indicated that following the Guidance Note is what matters, not the mere fact that AI was involved.
What remains unclear
The sources do not specify how the Commission will detect undisclosed AI use, nor whether the rules apply to documents filed before 20 October 2026 but heard after that date. The Guidance Note’s exact definition of generative AI is also not quoted in the available reporting, so it is unclear whether predictive text or grammar tools beyond the stated spelling and grammar carve-out are caught. Employers with matters already on foot should confirm transitional arrangements directly with the Commission or their legal advisers.
Sources: HRD America
Photo: Courtroom One Gavel - Flickr - Joe Gratz by Joe Gratz, CC0
