In Fernando Oliveira v Ryanair DAC (ADJ-00055225), the Complainant had sought €170,000 in compensation from his employer on grounds of numerous allegations including harassment, victimisation, unfair disciplinary actions, breach of contract, loss of earnings, and health and safety violations.
The Respondent categorically denied these allegations, and during the hearing raised numerous concerns in respect of the Complainant’s submissions which they alleged had been generated with the assistance of AI.
In particular, the Complainant’s submissions cited cases that: did not contain the outcome on which the Complaint relied, referenced awards of compensation in respect of claims that had been dismissed and, referenced two ‘phantom’ determinations that did not exist.
The Complainant initially put up a spirited defence to the insinuation his submissions were prepared with the assistance of AI, describing this as “baseless” and “designed to distract from the merits” of his case. However, at a subsequent hearing he later acknowledged he ‘may’ have used AI to draft them.
The Adjudication Officer (‘AO’) noted they were not particularly concerned about use of AI when drafting submissions, but emphasised that the parties to proceedings have an obligation to ensure that the content of their submissions is relevant and accurate.
The Complainant’s submissions were rife with citations that were “not relevant, mis-quoted, and in many incidences, non-existent.” This was categorised as “egregious” and an “abuse of process” by the AO who noted that this resulted in both the Respondent and AO undertaking a time-consuming exercisein verifying the accuracy or otherwise of these citations.
It is worth noting that this case was not dismissed because of the inaccuracies in the submissions, but rather because the Complainant failed to establish a prima facie case of discrimination. As the Complainant had only submitted a claim under employment equality legislation, all other claims were also dismissed.
In proceedings before the Workplace Relations Commission, an AO has the power to dismiss a complaint where they are of the opinion that it is frivolous or vexatious. This power has not often been used. The WRC on 30 October 2025 followed the Oliveira case with a guidance note. It simply refers to how submitting inaccurate or misleading information based on AI tools (or otherwise) may have an adverse impact on a party’s case. Of course, a Complainant could still succeed in a claim notwithstanding their reliance on inaccurate or misleading submissions, where this does not reach the threshold of being a frivolous or vexatious claim and they have otherwise proven their case.
We expect this will be the first of many cases on this topic as reliance on AI tools becomes more prevalent.
If you have questions about this article, please reach out to your usual Littler contact.
