Angelika Hellweger highlights another case in which AI-generated research was relied upon incorrectly in a court of law. The message is clear: mislead the court and pay the costs.
The legal sector around the world is cracking down on the use of AI tools for legal research. We discussed a case that chastised legal representatives for this practice in the Qatar Financial Centre and Commercial Court in November last year.
Now another case has emerged from the Abu Dhabi Global Market Courts (ADGM). The case is Arabyads Holdings Limited v Gulrez Alam Marghoob Alam [2025] ADGMCFI 0032.
Justice Paul Heath KC awarded costs on the indemnity basis against the defendant’s solicitors for submitting a ‘prolix’ defence, which included reference to cases that did not exist, and provided the incorrect citations for some cases. Additionally, the defence relied on cases to support propositions of law for which they did not stand.
It is clear that AI tools are becoming more popular in legal practice. However, these tools do not circumvent the rigour with which lawyers are expected to act. An AI tool may be able to suggest cases, but such cases should always be read thoroughly. The AI tools cannot understand nuance, and sometimes their summaries are inaccurate. Lawyers must still read cases in full to determine their application to the facts at hand.
Failure to apply the usual standards of legal research can result in misleading the court, even if that was not the intention. This is a serious concern and the courts are getting the message across that they do not expect lawyers to rely on AI tools without applying legal research conducted by a legally-trained person.
Wasted costs
In Arabyads Holdings Limited v Gulrez Alam Marghoob Alam [2025] ADGMCFI 0032, Arabyads’ position was that, due to the inappropriate use of AI, the defence prepared by MIO Legal Consultants LLP (MIO) “contained numerous false legal authorities and was prepared in a way that strongly suggests the improper use of [AI] tools, contrary to professional standards and MIO’s duties to the Court and the parties”.
Counsel for Arabyads complained about the amount of time that they had to spend (and the consequential cost to their client) in reviewing a defence of that nature and searching for authorities cited in the defence, some of which did not exist.
On that basis, Arabyads was awarded indemnity costs against MIO.
Reckless conduct
Justice Paul Heath KC accepted that the legal representatives (MIO) did not intend to mislead the court. However, intention is irrelevant in the circumstances.
The way in which MIO chose to research for and prepare the defence was a deliberate choice. MIO’s failure to verify whatever legal research was undertaken through AI, was reckless and amounted to a breach of the Rules of Conduct.
Key takeaways
There have now been judgments relating to the misuse of AI in the US, the UK, Qatar, and Abu Dhabi. Judicial sentiment around the globe remains the same: it is unacceptable to rely on legal research that is not thoroughly checked and verified.
In the UK, lawyers are officers of the Court, and they owe a duty to the Court not to mislead it, and to ensure that litigation is conducted efficiently and economically. Forcing the other side to waste significant cost in searching for cases that do not exist clearly contravenes these standards.
Lawyers using AI tools for research purposes should start from the premise that all authorities and/or articles on a particular topic that are revealed by AI research may not necessarily be accurately summarised in the response, or indeed may not exist. Thorough and rigorous legal research needs to be applied.
