Federal Court: Rejects AI Hallucination Defamation Argument in Australia

• Australia's Federal Court granted an applicant leave to withdraw an admission in a defamation case, despite the respondent's opposition.
• The respondent argued the applicant's explanation for the admission was nonsensical and characteristic of an AI hallucination.
• The court dismissed the AI hallucination claim, deeming the applicant's explanation plausible and sufficient.
• The decision, `Finnegan v Kemner (No 3) [2026] FCA 1402`, emphasized serving the interests of justice and minimizing prejudice.
• The applicant was ordered to pay the respondent's costs associated with the application and the withdrawal of the admission.

The Federal Court ultimately rejected the respondent's assertion that the applicant's explanation for the initial admission was nonsensical or indicative of an artificial intelligence hallucination, finding it to be plausible and sufficient given the circumstances.

In a significant ruling, the Federal Court of Australia recently addressed a defamation claim that saw an unusual argument involving artificial intelligence. The case, `Finnegan v Kemner (No 3) [2026] FCA 1402`, originated from an email sent by the applicant on 8 April 2019. The applicant subsequently initiated defamation proceedings against the respondent, alleging that an online noticeboard post by the respondent, which claimed the applicant was "trying to be anonymous and failing" when sending the email, was defamatory.

During the legal proceedings, a critical procedural point arose. On 20 November 2025, the respondent filed a substituted defence, specifically in paragraph 4(c)(i), asserting a defence of justification by pleading that the applicant had anonymously dispatched the email using an external affairs email address. In response, the applicant, in paragraph 4(c)(i) of a substituted reply filed on 5 December 2025, simply admitted to this allegation. This admission became the focal point of a subsequent application by the applicant to amend his pleading.

On 21 September 2026, the initial day of trial, the applicant sought leave under r 16.53 of the Federal Court Rules 2011 (Cth) to amend paragraph 4(c)(i) of his substituted reply. The proposed amendment aimed to acknowledge that the email did not identify him by name, but simultaneously deny that he sent it anonymously, given that the sender was identified as "External Affairs," a business name he utilized. The respondent vigorously opposed this application, contending that the applicant's explanation for making the initial admission was illogical and exhibited characteristics of a "verbose hallucination" — a trait the respondent attributed to artificial intelligence that the applicant allegedly employed for nearly all his recent communications in the matter.

Despite the respondent's objections, the Federal Court granted the applicant leave to amend paragraph 4(c)(i) of his substituted reply. The court directed the applicant to file and serve an amended substituted reply, strictly adhering to the court's orders. Furthermore, the applicant was ordered to cover the respondent's application costs and any costs incurred as a direct consequence of the amendment to the reply or the withdrawal of the admission.

The Federal Court ultimately rejected the respondent's assertion that the applicant's explanation for the initial admission was nonsensical or indicative of an artificial intelligence hallucination, finding it to be plausible and sufficient given the circumstances. While acknowledging that the applicant's explanation could have been more comprehensive, particularly regarding when the mistake was discovered, the court deemed it adequate and reflective of the relatively minor nature of the withdrawal. This ruling underscores a critical assessment of `Australia AI legal argument` when used as an explanation for procedural errors.

The court's decision to grant leave was rooted in the principle of serving the interests of the administration of justice. The judge noted that allowing a modest change would not jeopardize public confidence in the legal system, would not be unjust to the respondent, and would facilitate the determination of the actual matters in controversy. The court attributed the initial admission to inadvertence, highlighting that the question of whether the email was sent anonymously might still be a relevant issue in the proceeding. Despite the lateness of the withdrawal and amendment, the Federal Court perceived minimal prejudice to the respondent's conduct of the case, even though the respondent's counsel indicated a need to adjust their case strategy, opening and closing submissions, and cross-examination approach. The amendment, the court explained, would align with the applicant's evidence-in-chief and would only affect a dispute concerning the characterisation of facts related to the email, a characterisation the parties had previously agreed upon in a joint statement of legal and factual issues filed on 28 July 2026, and in other parts of the pleadings. This outcome provides important guidance on `artificial intelligence legal dispute Australia` and the standard for a `plausible explanation legal admission`.

Lawyers should note that Australian courts will critically assess claims of 'AI hallucination' as an explanation for errors or admissions, requiring plausible, human-attributable reasons rather than simply blaming AI. This case also reinforces the Federal Court's discretion to grant leave for late amendments to pleadings in the interest of justice, even when an AI-related argument is raised to oppose it.

Source: Original reporting via legal news outlet