The potential of artificial intelligence (AI) to transform the world of work, as well as its risks and dangers, has been widely reported.
But AI is now clogging up industrial relations courts.
AI can give workers a misplaced sense of confidence that there is a remedy for their workplace concerns, including for alleged bullying, adverse action or unfair dismissal.
Commissioners and magistrates have repeatedly expressed concern about AI-prepared submissions from workers who do not have union or legal representation.
Irrelevant references
The Fair Work Commission (FWC) recently threw out an AI-generated unfair dismissal case.
Deputy President Peter O’Keeffe said the applicant had cited three decisions, two of which were irrelevant and one that “simply did not exist”.
O’Keeffe had earlier advised the worker involved to be cautious when using AI.
The worker admitted to using AI in the submission, declaring it was accurate “to the best of my knowledge”.
However, O’Keeffe said: “Regrettably, the declaration that [the worker] had checked the material proved to be misleading”. He added that the worker had not read the cases, at least to “any meaningful extent”.
Dangerous and unpredictable
Federal Court judge Catherine Symons in July found that a delivery driver alleging adverse action had relied on “sycophantic” AI to submit their claim.
The judge said the case served as “a cautionary tale about the dangers of befriending AI-powered chatbots who masquerade as legal advisers”.
The worker had referred to their chatbot as an “industry expert”.
The worker was unable to produce the documents they had argued their case depended upon, and the matter was thrown out. Symons said the worker had come close to contempt of court for this.
Symons noted: “AI is not a lawyer, does not hold a practising certificate and is not authorised in any jurisdiction in Australia to engage in legal practice and to provide legal advice.”
She observed that AI does not care about the truth of the material it generates, and it can be dangerous and unpredictable without careful oversight and verification.
Obvious danger of relying on AI
The FWC in March heard a matter in which a worker was dismissed after sending numerous AI-generated emails to his employer about a bullying complaint the employer believed had been resolved.
Deputy President Tony Slevin referred to the “barrage” of emails as “dense, repetitive and often rambling” and said AI had ultimately led to the case’s failure.
Slevin said the employer had a valid reason to dismiss the worker, stating the employment relationship was “untenable owing to the way he dealt with the company in the communications and representations he made”.
Slevin had already thrown out another unfair dismissal application in 2025 noting “the obvious danger of relying on artificial intelligence for legal advice”.
He noted that ChatGPT had advised the worker to consult a legal professional or union representative. The worker did not take this advice.
No substitute for union support
In the examples here, workers were not represented by a legal professional.
The union believes AI is not an effective substitute for union advice or legal representation. Support a colleague today and invite them to join the IEU.