Startling warning: AI-generated fake judgements are shaking the pillars of justice in India.
India’s Supreme Court has escalated its response after discovering that a junior judge ruled on a property dispute using artificial-intelligence–produced, fake judgments. The apex court will review the lower court’s ruling from Vijayawada, in Andhra Pradesh, and the matter has been framed as an issue of grave “institutional concern” because the integrity of the adjudicatory process is at stake when AI fabrications influence real-world decisions.
This incident is part of a growing global conversation about how AI is altering court proceedings. The trouble surfaced last August, when a junior civil judge in the Vijayawada trial court issued an order in a property-dispute case. An official survey report had been requested, which the defendants objected to, and the judge dismissed their objection by citing four prior rulings. All four cited authorities were later found to be generated by AI, not genuine precedents.
While AI tools have transformed many workplace tasks, generative AI is notorious for “hallucinating”—presenting false information as fact and, at times, fabricating sources to back it up.
The defendants challenged the trial court’s order in the Andhra Pradesh High Court, pointing out the fake citations. The high court acknowledged the error but deemed it one of “good faith” and upheld the trial court’s decision nonetheless. In its ruling, the high court suggested that even if citations were fictitious, the core legal principles and their correct application could justify the order, effectively saying that incorrect references alone should not overturn it.
The high court asked the junior judge for an explanation. She testified that she was using AI for the first time and believed the citations to be authentic. She claimed no intent to misquote or misrepresent the rulings and attributed the error to reliance on an automated source, a point the high court accepted.
In a notable stance, the high court urged a shift from “artificial intelligence” to “actual intelligence,” signaling concern about overreliance on AI tools.
Following this, the defendants escalated the matter to the Supreme Court, which adopted a sterner posture toward AI’s role in judicial outcomes. Last week, the Supreme Court placed a stay on the lower court’s property-dispute order and asserted that using AI in judgments signifies misconduct, not merely a decision-making mistake. The court emphasized that this case concerns the process of adjudication itself, not just the merits.
The Supreme Court said it would conduct a deeper review and issued notices to India’s Attorney General, Solicitor General, and the Bar Council of India.
Separately, the court has expressed concern about lawyers drafting petitions with AI, calling such practices “absolutely uncalled for.”
This issue is not unique to India. In the United States, two federal judges faced criticism in October for AI-assisted rulings containing errors, while authorities in England and Wales cautioned lawyers in June 2025 against relying on AI-generated case materials after multiple instances of fictitious or altered rulings surfaced.
Across borders, legal institutions are grappling with how to regulate and monitor AI use in courtrooms. In response, India’s Supreme Court released a white paper on AI in the judiciary last year, outlining best practices and guidelines for AI use by judges, lawyers, and staff, while underscoring the necessity of human oversight and robust institutional safeguards.
