In a significant ruling that addresses the growing influence of Artificial Intelligence (AI) in the legal profession, the Andhra Pradesh High Court has held that a judicial order will not be set aside merely because it relies on non-existent citations generated by AI tools, provided the underlying legal principles applied are correct and in accordance with the law.
The court observed that while AI "hallucinations"—where software generates fake or fictitious information—pose a serious risk to the integrity of the justice system, the focus of a superior court should remain on whether the final decision aligns with established legal standards.
The matter arose from a civil suit (O.S. No. 773 of 2019) pending before the V Additional Junior Civil Judge in Vijayawada. The dispute involved a claim for a permanent injunction regarding a property. During the proceedings, the High Court had previously directed the appointment of an Advocate Commissioner to survey the land and identify if it matched the property descriptions in a specific 1989 sale deed.
Following the survey, the Advocate Commissioner submitted a report in July 2025. However, the defendants (petitioners) were dissatisfied. They filed an application (I.A. No. 457 of 2025) under Section 151 of the Code of Civil Procedure (CPC), asking the Trial Court to "strike down" the Commissioner's report. They alleged that the Commissioner had colluded with the plaintiffs, failed to take help from the government surveyor as instructed, and acted beyond the scope of the court's orders.
The Trial Court dismissed the defendants' application, stating that a Commissioner’s report is merely a piece of evidence that can be challenged during the trial and does not need to be "struck down" at an interim stage.
While this logic was sound, the Trial Court judge cited several decisions of the Supreme Court to support the order. When the defendants moved the High Court in a revision petition, their counsel, Sri M. Venkata Siva Teja, made a startling revelation: the citations used by the Trial Court judge did not exist.
Cases with names like "Subramani v. M. Natarajan" and "Lakshmi Devi v. K. Prabha," which appeared in the Trial Court’s order, were found to be fictitious. They could not be traced in any official law journal or legal database.
Taking note of this serious allegation, the High Court called for a report from the Trial Court judge. In an honest and regretful response, the Judicial Officer admitted that she had used an AI tool for research. She explained that she had acted in "good faith," believing the tool’s output to be accurate, and expressed deep regret for the oversight, promising to be more vigilant in the future.
Justice Ravi Nath Tilhari, presiding over the case, accepted the officer’s explanation but used the judgment to set clear boundaries for the use of technology in courts.
The High Court noted that AI, in its current state, is a tool for "organizing information" rather than "reasoning." The Court emphasized that judicial decision-making requires a "moral reasoning" and an "appreciation of human evidence" that machines simply do not possess.
However, on the question of whether the Trial Court's order should be cancelled because of these fake citations, the High Court adopted a pragmatic approach.