Artificial intelligence can significantly improve the efficiency and accessibility of India’s justice delivery system, but it cannot be allowed to replace judicial reasoning or human judgment, Justice Dinesh Kumar Singh, Judge, High Court of Karnataka, said at the ETLegalWorld’s AI-Powered Legal Transformation Summit 2026.
Delivering the keynote address on “The Rule of Law in the Age of Artificial Intelligence: Balancing Innovation, Justice and Accountability,” Justice Singh said AI is transforming not only businesses and everyday life but also the functioning of the legal ecosystem.
The challenge for the judiciary, he noted, is not to choose between tradition and technology, but to ensure that technological advancement remains aligned with the principles of justice.
“The rule of law is not merely a legal doctrine; it is a moral architecture of our constitutional order,” he said, emphasising that while technologies and algorithms will continue to evolve, constitutional values such as liberty, equality, dignity and justice must remain constant.
AI can improve access and judicial efficiency
Justice Singh said the judiciary has consistently adopted technology when it has helped strengthen access to justice.The transition from handwritten records to digital files, physical libraries to online legal databases and conventional courtrooms to virtual hearings has already demonstrated how technology can improve judicial administration without compromising judicial independence. AI, he said, is the next stage of that digital transformation.
Applications such as automated transcription, multilingual translation, intelligent case management and AI-assisted legal research can help courts reduce administrative burden and improve access to judicial services.
‘Justice is not a computational output’
Despite the opportunities, Justice Singh cautioned against viewing adjudication as a process that can be automated.AI may be effective at processing large volumes of information, but judicial decision-making requires understanding context, evaluating credibility, balancing competing rights and applying constitutional principles to unique factual situations.
“Justice is not a computational output; it is a reasoned human determination guided by law,” he said.
AI may assist the judicial process, but the ultimate authority to decide disputes involving life, liberty and dignity must remain with an independent human judge, he added.
AI hallucinations pose a serious legal risk
Justice Singh also highlighted the risk of AI-generated misinformation, particularly fabricated legal citations and judgments.
Sharing an example from his own experience, he said an AI system had produced several judicial precedents that appeared credible, but on verification, the cited judgments did not exist.
Such instances, commonly described as AI hallucinations, could have serious consequences in the legal system, where reliance on inaccurate authorities can affect legal arguments and judicial outcomes.
“The most dangerous misconception in the age of AI is to equate fluency with correctness,” he said.
Justice Singh stressed that AI-generated legal research should therefore be treated only as a starting point. Every output must be independently verified against authoritative sources, with the final responsibility for interpretation resting with judges.
Beyond accuracy, Justice Singh identified bias, explainability and transparency as key concerns in the deployment of AI within the justice delivery system.
He said constitutional guarantees of equality before law cannot permit discrimination to remain hidden behind the appearance of technological neutrality. Any AI system used by judicial institutions must therefore meet strong standards of transparency, explainability and fairness, he said.
AI could enhance legal research
Justice Singh pointed to legal research as one of the areas where AI could deliver significant gains.
As the volume of statutes, judgments and legal material continues to grow, conventional research is becoming increasingly time-intensive. AI-assisted systems can help analyse legal queries contextually, identify relevant principles and precedents and present structured insights more quickly. Such capabilities can allow judges and legal professionals to engage more deeply with complex or precedent-heavy matters.
However, he cautioned that AI systems remain dependent on the data on which they are trained and may generate incomplete, biased or inaccurate results.
AI may provide a persuasive answer, he said, but it does not independently understand the law or possess the ability to reliably distinguish between settled legal principles and errors. For Justice Singh, the debate is no longer about whether AI will become part of the legal system. “The journey has already begun,” he said.
Speaking at the ET AI-Powered Legal Transformation Summit 2026, Justice Singh concluded that the future belongs neither to institutions that resist technology nor to those that embrace it without safeguards, but to those that combine innovation with responsibility, wisdom and constitutional fidelity.
Today, the AI-Powered Legal Transformation Summit 2026 by ETLegalWorld has brought together general counsel, in-house legal leaders, and technology innovators, in Bengaluru, to chart the next phase of legal transformation in an AI-driven world.
