The Abu Dhabi Global Market (ADGM) Court has issued a detailed costs judgment that doubles as a practical guide for lawyers on the responsible use of artificial intelligence (AI) in litigation.
In a case arising from a dispute between Arabyads Holding Ltd (the Claimant) and former employee Gulrez Alam (the Defendant) (Case reference [2025] ADGMCFI 0032), the court addressed multiple costs applications and, crucially, sanctioned the Defendant’s former legal representatives for AI-related research failures that misled and burdened the Claimant and the court.
In this case, the legal representative of the Defendant prepared a defense spanning 327 paragraphs and 233 pages of exhibits, which required extensive review by the Claimant’s legal representative solely based on the number of pages submitted. In addition, the Claimant’s legal representative discovered that authorities within the pleading were fictious, wrongly cited, or not supportive of the proposition advanced.
In its judgment, among other points, the ADGM Court found that:
- AI research has increased in popularity among legal practitioners. However, it remains a legal practitioner’s obligation to verify any research prepared by an AI tool. This includes verification that the cited authorities exist and support the position that is advanced in the pleading. Without verifying any research, legal practitioners risk misleading the court by their submission. Failures in this case were described as “reckless,” amounted to a breach of ADGM Court Rules of Conduct 2016, and crossed the threshold for wasted costs.
- Despite the legal representative’s explanations about time, fee constraints, and the absence of UK counsel, the court found that lawyers remain bound to verify research and should withdraw if they cannot competently act. Filing a defective pleading without proper verification was a breach of duties owed to the court.
- The conduct of the legal representative was described as “unreasonable, if not improper” and they were ordered to pay AED 282,508 to the Claimant in respect of the costs for considering the defense, for preparing the application hearing in relation to the wasted costs and attendance at the hearing.
This judgment sets a clear benchmark for the ADGM and follows a judgment issued in 2025 by the Qatar Financial Centre Civil and Commercial Court which addressed the misuse of AI and the citation of non-existent cases. The judgment of the Qatar Financial Centre Court related to a citation by a lawyer that was described as a decision by the Qatar Financial Centre Court, but which did not exist (i.e., a “fake case”).
The rise in AI in disputes is being seen globally, and several cases have considered the inappropriate use of AI, including the following:
- A case in the US District Court for Southern District of New York in 2023, where lawyers used generative AI for legal research which produced a number of cases that did not exist.
- In 2025, the England and Wales High Court had to examine the professional duty of those using AI in research.
AI can be a useful tool, but unverified outputs are a professional hazard. It therefore remains the obligation of anyone using AI for research purposes to ensure that the relevant cases exist, support the proposition advanced, and are correctly cited.