On 29 July 2026, the President of the Irish High Court issued Practice Direction HC142 on the Responsible Use of Generative Artificial Intelligence in Court Documents, to take effect on 1 September 2026. HC142 applies to all civil proceedings before the High Court and covers documents prepared on or after the operative date, including documents filed in proceedings already underway at that date.
Practice Direction HC142 imposes two operative requirements. First, every affidavit and witness statement filed after 1 September 2026 must carry an express declaration that generative AI was not used to generate the substantive content of the document. The direction defines the prohibition broadly, extending it to AI use that alters, embellishes, strengthens, dilutes, or rephrases the evidence contained in the document. Purely administrative uses of AI, including document formatting and scheduling tasks, do not engage the declaration requirement. Second, any practitioner who uses generative AI in any capacity in preparing a court document must independently verify every claim, citation, and legal reference before filing. The direction states that checking AI-generated content against a second AI tool does not constitute adequate verification.
Solicitors, barristers, expert witnesses, and self-represented litigants before the High Court are all within scope. Solicitors bear personal professional responsibility for the accuracy of all filed documents regardless of who prepared the content; that responsibility cannot be transferred to an AI system or delegated to a junior who used one. Experts producing court reports must append the same substantive-content declaration as lay deponents. Non-compliance exposes practitioners to court sanctions, professional disciplinary proceedings, and costs orders.
HC142 was issued in direct response to the Court of Appeal judgment in Guerin v O'Doherty [2026] IECA 48, in which the court noted a litigant's reliance on AI-generated material containing references to non-existent authorities. The direction warns that generative AI systems produce fictitious case citations, invented statutory provisions, factual errors, and biased summaries of genuine sources. No carve-out applies for AI systems marketed as legal research tools.
Licentium advises law firms, corporate legal teams, litigation funders, and expert witnesses on AI governance and responsible AI deployment in regulated professional environments. Work we undertake includes AI use policy drafting for law firms and regulated entities, compliance assessment under the EU AI Act, high-risk AI system classification, professional liability analysis for AI-assisted legal and expert work, and cross-border AI litigation risk assessment.