The Fair Work Commission is cracking down on AI-generated claims, and employers need to pay attention too

The short version

Over the past year, the Fair Work Commission and Federal Circuit and Family Court have dealt with a string of cases where applicants, and in some cases employers, submitted material prepared with generative AI that turned out to contain fabricated case citations, invented legal principles, or evidence that simply didn’t exist.

The Commission has now released a formal Guidance Note requiring disclosure of AI use in Commission matters, effective from 20 October 2026. If your business ever ends up defending a claim, or drafting one, this affects you directly.

What’s actually been happening

Commission members have reported identifying signs of generative AI use in a significant share of the matters they’ve handled in recent months. Several decisions this year have called it out directly and not gently.

• In one unfair dismissal appeal, the Full Bench criticised an applicant’s reliance on AI-prepared submissions and communications, noting that generative AI “is merely a tool” and “not capable of making the type of decisions required to effectively conduct litigation.”
• In another matter, an applicant cited legal authorities that simply didn’t exist, described by the Commission as AI hallucinations with no actual legal basis.
• In a general protections matter before the Federal Circuit and Family Court, a worker’s evidence list was found to have been AI-generated, referring to documents that either didn’t exist or couldn’t be identified. The court found his continued reliance on it came “perilously close” to contempt, and dismissed the claim in full.
• In another Commission case, a worker relied on AI-generated contract clauses and award provisions that turned out to be fabricated. The Commission didn’t just dismiss the claim, it invited the employer to apply for costs.

What’s changing from 20 October 2026

The Commission’s new Guidance Note on the use of generative AI applies to all individual rights matters, including unfair dismissal and general protections applications. Anyone using tools like ChatGPT, Claude, Copilot or Gemini to prepare an application, submission or witness statement will need to disclose that use and verify the accuracy of what’s produced before it’s filed.

Importantly, the Commission hasn’t framed this as a blanket warning against using AI. The guidance leans toward encouraging responsible, verified use rather than discouraging it outright, with practical prompting guidance included alongside the disclosure requirement.

Why this matters for employers, not just applicants

It’s tempting to read these stories as an employee problem. It isn’t. Employers and their representatives are just as capable of filing a response built on an AI tool’s confident but wrong output, and the Commission has been explicit that a response relying on incorrect AI-generated content, missed procedural steps, or case law that doesn’t exist can weaken an otherwise solid defence.

There’s also a practical flip side: if you’re on the receiving end of a claim that reads oddly, repeats the same points with shifting reasoning, or cites cases you can’t find anywhere, that’s now a recognised pattern the Commission is actively looking for, and it’s worth flagging rather than assuming it’s just unusual drafting.

What this means for your business

• If you or anyone drafting responses on your behalf uses AI tools, verify every citation, every legal principle and every factual claim before it goes anywhere near the Commission. Don’t take fluent output at face value.
• From 20 October 2026, be ready to disclose AI use in any submission, application or witness statement you file with the Commission.
• Don’t assume AI-assisted material is a shortcut. The evidence so far shows unverified AI use is more likely to produce weaker, easily unpicked claims, whichever side is relying on it.
• If you’re managing a claim internally without legal support, get it checked by someone who knows the actual case law before you rely on anything a chatbot has handed you.

Frequently asked questions

Do I have to disclose if I used AI to prepare a Fair Work Commission submission?

From 20 October 2026, yes. The Commission’s Guidance Note on generative AI requires anyone who uses tools such as ChatGPT, Claude, Copilot or Gemini to prepare an application, submission or witness statement to disclose that use, and to verify the accuracy of the resulting document before filing.

What happens if a submission contains AI-generated fake case citations?

The Commission has already dismissed claims and invited costs applications against parties who relied on fabricated case law or invented evidence. In at least one instance before the Federal Circuit and Family Court, continued reliance on AI-generated material came close to contempt of court. Fabricated citations are treated seriously, not as a minor drafting error.

Can employers use AI to help prepare their response to a claim?

Yes, and the Commission’s guidance doesn’t discourage it. The key requirement is verification: every case cited, every legal principle referenced and every factual claim needs to be checked against the real source before filing, and AI use will need to be disclosed under the new rules from October 2026.

Bottom line

Facing a Fair Work claim and want it handled properly, not by a chatbot? HR Gurus helps Australian businesses cut through the BS and respond to claims the right way. Get in touch with the team.