Generative AI: A Boon Or Bane In The Legal Landscape.

Introduction

With the inception of AI (Artificial Intelligence), there were numerous debates and deliberations about its consequences, such as job replacement or the transition towards advanced forms of technology. Putting all of that aside upon realizing its potential to offer amplified levels of productivity, efficiency, and reduced costs, the legal sector adopted generative AI to simplify the process of legal research, contract reviewing, document analysis, and case management with an eye on speed, efficiency, and accuracy for litigants, parties, and courts.

Having said that, AI is not infallible. There are very high chances of it making mistakes, because, at the end, the information is sourced by AI through a data feed, so it is not necessarily authentic or legitimate.

Recently, in the case of Arabyads Holding v. Gulrez Alam Marghoob Alam (2025) ADGMFCI0032, the Abu Dhabi Global Market (ADGM) Court of First Instance found that the Defendant’s legal representatives made use of legal documents that refer to authorities were misapplied, non-existent or mis-cited hence making them liable to pay approx. AED 282,508 to claimant as wasted costs, assessed on an indemnity basis, in relation to the defences associated costs as noted in the judgment.

Transfiguration of Legal Research

AI immensely moulded traditional means of legal research. Advanced legal research tools can now scrutinise and construe thousands of judgments ranging from various jurisdictions and subject matters within a few minutes or seconds. Predictive analytical tools aids litigators in understanding the plausibility of success derived from old case patterns; on the other side, document-reviewing algorithms enhances coherence in due diligence and disclosure exercises. Various judicial systems have acknowledged AI’s honest utility. Courts under the common law jurisdiction endorse AI-assisted document management and e-discovery in complex commercial litigation. At the same time, courts have reinstated that it should only be used under proper supervision, and warned that AI is merely an aid and cannot be substituted for professional ethics, judgment, and advocacy.

Downsides to AI in Legal Research

The confident generation of inaccurate, partial, and deceptive information, including distorted legal doctrines and non-existent case laws are the chief risks associated with generative AI. In contrast to the conventional databases, generative AI does not comprehend or distinguish between probable fiction and binding authority unless meticulously constrained and verified. Courts in various jurisdictions has alerted that over reliance on AI without proper surveillance amounts to professional misconduct.

The English High Court in R (Ayinde) v. London Borough of Haringey (2025) EWHC 1383and the Superior Court of Justice of Ontario in Ko v Li (2025) ONSC 2766, laid down that advocates who take the help of AI should verify the output against a legit authority and that failing to do so is gross violation to the administration of justice.

Case Study on Arabyads Holding v. Gulrez Alam Marghoob Alam (UAE, 2025)

The landmark decision rendered by the Abu Dhabi Global Market Court in the case Arabyads Holding v. Gulrez Alam Marghoob Alam (2025) ADGMFCI0032, reflects the timely intervention in the landscape of AI assisted legal research in the UAE. Factual matrix of the case, the defendant’s legal representative filed written pleadings citing various legal authorities. Upon inspection, the Court identified that several cases were non-existent and mis-cited. Justice Paul Heath KC observed that such fallacies are the epitome of unverified AI generated legal research. The Court held that advocates owe a duty to the Court to guarantee that the information used is relevant, authentic, and legit. The result being the Court of First Instance imposed the wasted cost of AED 282,508/- on an indemnity basis, emphasizing that the callous use of AI invites both compensatory and punitive consequences. The decision aligns ADGM jurisprudence in conformity with the common law jurisprudence and points out that technological assistance cannot weaken professional duties.

Comparative Study Across Various Jurisdictions

Courts in the United States have, to a great extent, used the Federal Rule of Civil Procedure, Rule 11, to sanction lawyers for fake citations, as in Mata v. Aviance Inc., 678 F. Supp.3d 443 (S.D.N.Y. 2023), wherein Justice P. Kevin Castel sanctioned lawyers $ 5,000 for AI -generated (ChatGPT) brief. Similarly, in California’s Noland v. Land of the Free, L.P. (2025) 114 Cal.App.5th 426, Justice sanctioned the lawyers $10.000 for using generative AI for filing, and their briefs contains fabricated legal citations. In the English High Court, California’s Noland v. Land of the Free (2025) warned about the excessive unverified usage of AI.

Conclusion

Unquestionably AI play very transformative role in the field of legal practise and research. If utilised with judicious care, it can promote efficiency, availability, and experimental depth. Nonetheless, the jurisprudence arising around the globe, and concluding in Arabyads Holding v. Gulrez Alam Marghoob Alam makes it crystal clear that it is the lawyer who bears the consequences of such hideous errors and not the AI or algorithm. Hence, ethical training, professional discipline, and judicial oversight play a significant role in the future of AI in the legal field. Innovation will continue to be entertained until it adheres to truthfulness and the rule of law.