Riley v. Nuvei: Fair Work Commission Notes AI Hallucinations in Self-Represented Applicant’s Case (Australia, Jan 2026)

Case at a Glance

Background

Mr. Jason Riley was dismissed from Nuvei Australia Merchant Services Pty Ltd on 19 June 2025 after a prolonged absence from work. He filed an unfair dismissal application under Section 394 of the Fair Work Act 2009. Riley appeared in person. During the proceedings, Riley disclosed to Commissioner Redford that he had obtained assistance in preparing his case from an AI model that was ‘legally trained’. The Commission noted that some of Riley’s legal references and principles may have lacked genuine legal basis.

The AI Issue

The Commission had to determine both the substantive merits of the unfair dismissal claim and how to treat the self-represented applicant’s reliance on a ‘legally trained’ AI model. Some of Riley’s cited legal principles appeared to be AI hallucinations — fabricated rules or cases with no basis in Australian employment law. The Commission noted that Riley conceded these points when they were queried, rather than pressing fabricated authorities.

• The unfair dismissal application was dismissed — Riley could not perform the inherent requirements of his job [substantive finding].
• Commissioner Redford noted Riley’s use of a ‘legally trained’ AI model in the published decision [AI disclosure on record].
• Some of Riley’s legal references were identified as possible AI hallucinations [fabricated legal principles].
• Riley’s credit was preserved by the fact that he conceded the hallucinated references when challenged [mitigation by disclosure].
• The Fair Work Commission is the first Australian employment tribunal to formally note AI assistance in a published decision [landmark for employment law].

“Mr Riley made clear that he had obtained assistance in the presentation of his case by an Artificial Intelligence (AI) model, which was ‘legally trained’. Some of the references made by Mr Riley to particular legal principles or authorities may have been AI hallucinations and lacked any legal basis. To Mr Riley’s credit, he did not press these when they were queried.”

— Commissioner Redford, Fair Work Commission, 19 January 2026

The India Angle

Indian Law Equivalent: India’s labour tribunals — the Industrial Tribunal, Central Government Industrial Tribunal (CGIT), and Labour Courts — regularly deal with self-represented workmen in unfair dismissal (retrenchment) disputes. Under Section 25F and 25G of the Industrial Disputes Act, 1947, similar claims are filed. As AI tools become accessible in India, self-represented workers may use them to prepare submissions. The Riley framework — note the AI use, be lenient if the party concedes errors — provides a useful template for Indian labour tribunals.

Bar Council Rules: BCI Rules apply to enrolled advocates, not self-represented litigants. However, if an advocate assists a party using AI and submits hallucinated legal references to a labour tribunal, BCI Rule 15 is engaged. The Indian advocate’s duty to verify is unaffected by the fact that the forum is a tribunal rather than a court.

Practical Advice for Indian Advocates: When representing or advising self-represented parties who have used AI tools to prepare their cases, always review and sanitize AI-generated content before submission. An advocate who files an AI-hallucinated reference — even on behalf of a pro se litigant — takes professional responsibility for that citation.

Quick Takeaways

• Voluntary disclosure of AI use and conceding errors when challenged are mitigating factors in AI hallucination cases.
• Employment tribunals are now part of the AI hallucination landscape — not just civil courts.
• Indian labour tribunals will face self-represented AI-assisted litigants — courts should develop a consistent approach similar to the FWC’s.