High Court issues new Practice Direction on the use of AI in Irish legal proceedings

At a glance

• In May 2026, we reported on the Labour Court's guidance and the Court of Appeal's decision in James Guerin v Gemma O'Doherty (see our previous update), both of which addressed the responsible use of AI in legal proceedings.
• The High Court has now issued Practice Direction (HC142), introducing the most comprehensive Irish framework to date governing the use of generative AI in court documents.
• While recognising the potential benefits of AI, the Practice Direction emphasises that responsibility for the accuracy of submissions, evidence and legal authorities remains with the user, and introduces mandatory declarations for certain categories of evidence.

A new framework for the use of AI in Litigation

The Practice Direction takes effect on 1 September 2026 and applies to all civil proceedings before the High Court, including proceedings already before the High Court on that date.

The Practice Direction acknowledges that AI tools can improve efficiency, reduce costs and enhance access to justice. However, it makes clear that parties, legal practitioners, witnesses and experts remain fully responsible for the accuracy and integrity of any material submitted to the High Court.

The use of AI does not alter existing professional, ethical or procedural obligations.

Human verification remains essential

A central theme of the Practice Direction is the need for meaningful human oversight. Any content generated with the assistance of AI must be independently verified before it is used in proceedings. This includes checking facts, legal authorities, quotations and references against authoritative sources. Significantly, the High Court states that verifying AI-generated content using another AI tool is not sufficient.

The Practice Direction also highlights the well-known risks associated with generative AI, including 'hallucinations', fabricated legal authorities, factual inaccuracies, bias, confidentiality concerns and potential privilege issues.

New safeguards for evidence

The most significant requirements relate to witness evidence and expert reports.

Witness statements and affidavits must generally include a declaration confirming that generative AI has not been used to generate the substantive content of the evidence and that the document reflects the witness' own knowledge and recollection. A template declaration is provided in the Practice Direction.

AI may be used for administrative and ancillary tasks, such as formatting and spell-checking. However, these uses must still be addressed through the prescribed declaration, which makes clear that AI has not been used to generate or modify the substantive evidence of the witness or deponent.

Similar safeguards apply to expert reports. Experts must confirm that the substantive content of their report reflects their own opinions and expertise and, where AI has been used, disclose the nature and extent of that use where necessary to allow the High Court and parties to assess the reliability and independence of the opinion evidence.

Disclosure and sanctions

Although there is no general obligation to disclose the use of AI in preparing court documents, the High Court may require parties to explain whether AI tools were used and the nature and extent of that use. Parties and practitioners should therefore ensure they are in a position to provide this information if requested.

Failure to comply may have serious consequences, including adverse cost orders, rejection of court documents, strike-out of claims or defences, and, in the case of legal practitioners, referral to the Legal Services Regulatory Authority or their professional body.

Key takeaways for employers

• AI can assist, but responsibility remains with the user. Any AI-generated content used in investigations, disciplinary processes or litigation should be carefully reviewed and independently verified before being relied upon.
• Review AI governance and oversight measures. The Irish courts are placing increasing emphasis on accountability, transparency and human oversight in the use of AI. Employers should ensure appropriate controls are in place around the use of AI in HR and legal processes.
• Exercise caution when preparing evidence. The Practice Direction introduces specific requirements for witness statements, affidavits and expert reports, reinforcing the importance of human authorship, review and responsibility in employment disputes and litigation generally.