AI can be an efficient tool in the legal profession but needs rigorous monitoring, say legal experts

Artificial intelligence is increasingly finding a place in the legal profession, but its growing use has also sharpened concerns over accuracy, accountability, and the limits of machine reasoning.

At a panel discussion on the impact of artificial intelligence (AI) in the legal profession at Justice Unplugged, a legal conclave jointly hosted by the VIT School of Law and TheHindu, practitioners and academics described AI as a powerful aid but one that requires rigorous human oversight.

Supreme Court Advocate-on-Record Neha Rathi said AI was “already acting as a great efficient tool” in legal practices. From summarising lengthy briefs and generating concise points to assisting with translations, AI tools and specialised legal software are now routinely used by lawyers, particularly for urgent matters, Ms. Rathi noted.

Due diligence is key

“It is a very efficient intern or a junior,” she said, pointing out that AI can quickly shortlist relevant judgments and help structure arguments.

However, Ms. Rathi cautioned against a blind reliance on AI, pointing to instances where fictitious case citations have surfaced in court filings. Recalling a recent episode where Justice B.V. Nagarathna of the Supreme Court flagged a non-existent judgment, Mercy versus Mankind, she underscored the need for meticulous verification. 

“Due diligence is very, very important. We cannot rely on it with our eyes closed,” she said. “Every word has to be the same as the judge wrote in a particular judgment. There is no running away from that verification,” she added.

Tool for transcription, case management

Advocate Vishal Sinha, who practices in the Supreme Court, described AI as “a great tool” whose impact could be compared to the advent of the internet. “We are now realising that there is this huge repository of tools which are being modulated into different fields of life — law being one of them,” he said. 

He noted that courts have already integrated AI-driven tools such as live transcription during Constitution Bench hearings, document translation, and case categorisation.

On the question of pendency, Mr. Sinha said AI could assist by summarising voluminous records and streamlining case management. “It is far more reliable with known data than asking questions with respect to the unknown,” he said, adding that matters requiring moral reasoning should remain within the domain of judges.

Cannot replace humans

Speaking from an academic perspective, VIT School of Law professor Jishnu J.R. said that AI improves efficiency “with respect to quantity,” but cautioned that “with respect to quality, we will have to look into it.” 

“At VIT School of Law, we introduced Artificial Intelligence and Law into the syllabus three semesters ago. While the technical component may not be very deep — since we are a law school — we are bridging that gap by inviting industry experts and cybersecurity professionals,” Dr. Jishnu added. While data-driven tools may aid evidence-based determinations, he argued that areas involving moral conscience and contextual sensitivity cannot yet be delegated to machines.

The panel, moderated by Sandeep Phukan, Associate Editor, TheHindu, agreed that AI may enhance productivity, but it cannot be a substitute for the human element central to the administration of justice.