WRC Warns Against AI-generated Written Submissions

Our Employment Law & Benefits team reviews a recent Workplace Relations Commission case where the Adjudication Officer was critical of the complainant’s suspected use of artificial intelligence (AI) in preparing his submissions. The team explores why the case serves as an important reminder of the requirement for relevant and accurate legal submissions, and the risks associated with the use of AI in drafting generally.

What you need to know

• Scope of the claim: The complainant alleged that he was discriminated against on the grounds of family status and race by his previous employer, Ryanair. He also alleged victimisation, harassment, sexual harassment and procedural unfairness in Ryanair’s disciplinary procedure.
• Unsuccessful result: Each of the complainant’s claims were unsuccessful before the Workplace Relations Commission.
• Issues with complainant’s submission: The complainant’s submissions included citations that were not relevant, were misquoted or, in certain cases, were non-existent.
• Use of AI: Ryanair suggested that the complainant used AI in drafting his submissions and the complainant rejected this as “baseless”. Subsequently, however, he acknowledged that he may have used AI in preparing his submissions.
• Adjudication Officer warning: The Adjudication Officer warned that parties have an obligation to ensure that their submissions are relevant and accurate and do not set out to mislead.

The claim

The recent case of Fernando Oliveira v Ryanair DAC[1] serves as a stark reminder of the importance of relevant and accurate legal submissions and the dangers associated with overreliance on AI in drafting.

A Ryanair cabin crew worker brought a number of claims against Ryanair, alleging that he was discriminated against on the grounds of family status and race regarding his promotion, provision of training and conditions of employment. He also made an allegation of sexual harassment. He further alleged that he was victimised and harassed on foot of false allegations made by a colleague against him and that he was subject to unfair disciplinary actions.

Ryanair rejected the claims, arguing that they were frivolous, vexatious and bound to fail.

Use of AI

Ryanair suggested that the complainant’s submissions “may have been generated with the assistance of artificial intelligence”. This was particularly evident where the case citations relied upon did not appear to give the outcome which the complainant alluded to. A specific example of this was the complainant’s reliance on citation: DEC-E2016-105. The complainant asserted that this case related to racial harassment. In reality, however, it was a claim for discriminatory dismissal and made no reference to harassment save by reference to another authority cited in the decision. The complainant asserted that €15,000 was awarded in that case, however, in reality, the case was dismissed.

At least two other of the cases referred to by the complainant appeared to be AI hallucinations. When this was raised by Ryanair, the complainant characterised the suggestion that he had used AI in drafting his submissions as “baseless”.

Adjudication Officer warning

The Adjudication Officer determined that none of the complainant’s complaints were well-founded. The Adjudication Officer stated that on initial inspection, it appeared that the complainant’s submissions had been a representative of the complainant, but closer inspection revealed this not to be the case. She noted that, initially the complainant rejected any use of AI in the drafting of his submissions, but that on day two of the hearing, he acknowledged that he may have used it.

The Adjudication Officer stated that she was not particularly concerned about whether the complainant used AI in drafting the submissions or not. However, she warned that both parties had an obligation “to ensure that their submissions are relevant and accurate and do not set out to mislead either the other party or the Adjudication Officer.”

The Adjudication Officer went on to criticise the complainant’s submissions for being “rife with citations that were not relevant, mis-quoted and in many instances, non-existent.” She noted that a considerable amount of Ryanair’s and the Adjudication Officer’s time was “wasted” trying to establish whether the citations provided by the complainant were legitimate or not.

She added that the complainant’s attempts to rely on hallucinated citations to support his claims, “can only be described as egregious and an abuse of process”.

Key takeaway

Although AI can be an effective tool for structural drafting and legal research, it is vital that a qualified legal professional critically review any AI-generated content to ensure accuracy and relevance. For expert drafting assistance, please contact our Employment Law & Benefits Team.

The content of this article is provided for information purposes only and does not constitute legal or other advice.

[1] ADJ-00055225.