AI is being used all over the world in healthcare, medicine, business, engineering, education, law enforcement, military and even in various governmental administrative decision-making processes. The judiciary is no longer lagging behind in this regard. Judges, court staff, lawyers, and self-represented litigants are also applying AI tools for numerous needs and using it in legal proceedings. AI tools have potentials to modernize the judicial proceedings and automating administrative tasks.
Some judges have already created discussion by applying popular AI chatbot like ChatGPT in judicial decision-making processes. Lawyers also often present references, arguments, or precedents prepared through AI in court that do not actually exist. Consequently, the regulatory bodies of lawyers and the judicial bodies in different countries have drawn boundaries to regulate the unwanted and unethical use of AI in the judicial proceedings. Because the unplanned, improper and unbridled application of AI systems in pleadings, arguments, evidence analysis or in the judgments can bring severe consequences. As a result, a crisis of trust among citizens about the justice system can be appeared and questions relating to the independence, transparency, accountability, legitimacy etc., regarding the judiciary may be arisen.
Recently, the Supreme Court of India (SCOI) noticed that cases had been determined by referring to non-existent and fictional precedents derived from AI. A junior civil judge from Vijayawada city of Andhra Pradesh took aid from AI chatbot to adjudicate a property related suit. Again, in Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. same incident was identified by SCOI.
Therefore, the court has warned about the legal consequences of such unintended use of AI in judicial decisions. The apex court held that the judgment has now become a cause of considerable institutional concern in India not because of the decision taken on the merits of the case, but because of the unethical use of AI in the trial and adjudication process. The highest court in India further opined that the judgment given based on a fictitious judgment will not only be seen as a judicial error but will also be considered as judicial misconduct.
To prevent the unwanted use of AI in judicial decisions, the SCOI has already formulated a draft guideline for judges. There is also a committee in the Supreme Court working on how AI can be used in the judicial systems in India. The SCOI has planned to start using AI technology in case listing, scheduling of hearings and bench allocations. There is a lack of transparency in case listing in our country too and irregularities are being seen in some places. Even allegations of corruption have come to the fore regarding case listing. In that context, specific and planned use of AI technology can bring transparency and reduce corruption.
A survey conducted by the Northwestern University and released on 30 March, revealed that more than 60 percent judges are using popular AI tools at least once in their judicial work. Among them, around 22 percent confirmed they used AI daily or weekly in their duties although they didn’t receive any training regarding the prospects and challenges of such AI systems from court administration.
Not only judges, but also lawyers are employing AI tools uncritically in court pleadings. In multiple occasions, information obtained from these tools are being exported directly to case documents without verifying the authenticity and as a result, repealed laws or overruled judgments, fictitious information are being incorporated into those documents. According to a global database in this regard, as of September 14, 2,041 cases have been identified worldwide where generative AI has created misleading or non-existent contents in court adjudication. A significant constraint of chatbots like ChatGPT is the random production of fictitious information often called as AI hallucination with great confidence and present them convincingly. Besides, AI chatbots can present biased, discriminatory, unrepresentative outputs. Moreover, AI lacks the fundamental human qualities required to conduct legal proceedings.
A Federal Court in West Virginia has recently issued an order to consider the use of AI by the defence to review sensitive discovery (e.g., witness info, medical records, footage involving minors). It mandates defence counsel to obtain the government’s written approval before inputting such materials into any AI tool. The counsel must ensure that no information will be stored for training purposes, or be shared with third parties, or be retained after a case is over. That reminds about the increasing judicial awareness of the potential of AI to affect confidentiality and the integrity of discovery.
Therefore, it is very important to know when and how AI should be used in court, along with its operational functions and challenges. AI tools may reduce human effort, cost or time in specific clerical or administrative tasks. However, if the own intelligence of lawyers or judges becomes dependent on AI, then the results will be dire, and the cost will be paid not only to the litigants but also to the entire population. AI can never be a substitute for the intelligence of a lawyer or judge. It can generate language with fluency although it cannot take over judgment.
Accordingly, several countries including the UK, Ireland, Australia, New Zealand, Pakistan, Hong Kong and more have already adopted and implemented guidelines or practice directions as to how AI systems can be applied in the legal profession. Although there is debate about how effective those guidelines are, and research on those debates is also ongoing. But in Bangladesh, the Bar Council, the regulatory body for lawyers, and the Supreme Court have not yet provided any guidelines or training for lawyers, judges and court staff to use AI. Although so far, no allegations have been revealed that anyone in the country has presented fictitious references obtained through AI in the court. But that certainly does not mean that lawyers and judges are not using AI tools in our country.
Although the initiative to formulate a national AI policy is underway; but it is urgent to adopt a separate policy specifically related to the use of AI for the judiciary. Hence, before the burden of judging is transferred to AI, or before a major disaster occurs in judicial sector, it is urgent for policymakers to actively think about practice directions or policies on the use of AI in courts now. At the same time, extensive research, discussion, allocation of budget and useful practical actions are obvious on how modern technology like AI can be used to prevent corruption, enhance the quality of services, reduce the suffering of litigants, ensure transparency and accountability, reduce the cost, complexity and delay of cases etc., while maintaining the basic principles of justice. We should always bear in mind that courts are not only the forum to resolve disputes; rather they uphold human rights, represent the rule of law and make democratic institutions accountable.
Published in Ajker Patrika as op-ed on 4 April 2026 on page 4.
Published in the Daily Asian Age as op-ed on 16 September 2026 on page 7.
No comments:
Post a Comment