AI-Generated Citations Cannot Replace Judicial Mind, But Order Will Stand If Law Applied Is Correct: Andhra Pradesh HC

The Andhra Pradesh High Court held that while unverified AI-generated citations pose serious risks to the justice delivery system, a judicial order will not be vitiated solely for such references if the correct legal principles have been independently applied

By :  Sakshi
Update: 2026-01-28 16:25 GMT

Use of AI Requires Caution, Not Automatic Reversal: Andhra Pradesh High Court Dismisses Revision Where Legal Principles Were Properly Applied

The Andhra Pradesh High Court has dismissed a Civil Revision Petition challenging a trial court order which had relied upon non-existent judicial precedents generated through an Artificial Intelligence tool, holding that an order cannot be set aside solely on this ground when the correct principles of law have otherwise been applied.

The revision petition was heard and decided by Justice Ravi Nath Tilhari, exercising supervisory jurisdiction under Article 227 of the Constitution of India.

The petitioners, who were defendants 1 and 2 in O.S. No. 773 of 2019 pending before the V Additional Junior Civil Judge, Vijayawada, had approached the High Court challenging an order dated 19 August 2025.

By the said order, the trial court had dismissed an application filed under Section 151 of the Code of Civil Procedure seeking to strike down the report of an Advocate Commissioner submitted pursuant to earlier directions of the High Court.

The original suit had been instituted by the respondents for permanent injunction and other consequential reliefs. During the pendency of the suit, the plaintiffs had moved an application seeking appointment of an Advocate Commissioner, which was initially dismissed. The defendants thereafter approached the High Court in an earlier Civil Revision Petition, which was allowed by an order dated 12.12.2024.

The High Court had directed appointment of an Advocate Commissioner to identify whether the plaint schedule property fell within the property purchased by the first defendant under Document No. 6404 of 1989 and had further directed that the assistance of the Town or Mandal Surveyor be taken for the purpose of the survey.

Pursuant to the said directions, the Advocate Commissioner conducted inspection and submitted a report dated 7 July 2025. The defendants subsequently filed I.A. No. 457 of 2025 contending that the report deserved to be struck down as the Advocate Commissioner had allegedly colluded with the plaintiffs and had failed to adhere to the High Court’s directions, particularly with respect to taking assistance from the Town or Mandal Surveyor.

The trial court framed three points for consideration, namely whether the application under Section 151 CPC was maintainable, whether the Advocate Commissioner had acted beyond the scope of the High Court’s directions, and whether allegations of collusion and misconduct had been substantiated. All three points were answered against the defendants, with the trial court holding that the commissioner’s report was only a piece of evidence to be assessed during trial and that allegations of bias or misconduct required proof by cogent evidence.

While dismissing the application, the trial court relied upon certain Supreme Court decisions which were later found to be non-existent.

Before the High Court, counsel for the petitioners argued that the impugned order was unsustainable as it was based upon fictitious rulings and citations generated by an AI tool. It was submitted that reliance upon non-existent authorities rendered the order legally vulnerable and liable to be set aside.

In view of this contention, the High Court had earlier called for a report from the trial court. In its report, the Judicial Officer acknowledged that the citations had been generated through an Artificial Intelligence tool and that she had relied upon them in good faith without independent verification. She clarified that the judgments could not be traced on official databases or law journals and expressed regret, assuring greater caution in the future. The High Court accepted the report.

Addressing the broader issue of use of Artificial Intelligence in judicial decision-making, the High Court observed that AI, in its present stage of development, could only assist in organizing information or summarizing material and lacked the capacity for moral reasoning or appreciation of evidence. The Court noted that unverified reliance on AI-generated material could lead to fabrication of authorities and misuse, thereby undermining the integrity of judicial proceedings.

However, on the core issue, the High Court declined to interfere with the trial court’s order.

It held that mere mention of non-existent citations would not vitiate an order if the principles of law applied were otherwise correct and in accordance with settled legal position. The Court clarified that interference would be warranted only where incorrect legal principles were applied or where the application of law was fundamentally flawed due to reliance on AI-generated content.

Referring to Order XXVI Rule 10 of the Code of Civil Procedure and authoritative precedents of the Supreme Court, the High Court reiterated that an Advocate Commissioner’s report is not binding and only aids the court in adjudication. Such a report forms part of the evidence and its weight is to be assessed during trial. Allegations of collusion or misconduct must be established through evidence, including cross-examination of the commissioner, and cannot be accepted on mere assertions at an interlocutory stage.

The Court held that the trial court had correctly applied the law governing the evidentiary value of commissioner reports and had preserved the defendants right to challenge the report during trial.

As no jurisdictional error or illegality was found, the Civil Revision Petition was dismissed, with no order as to costs. All pending interlocutory applications were also closed.

Case Title: Gummadi Usha Rani & Anr. v. Sure Mallikarjuna Rao & Anr.

Bench: Justice Ravi Nath Tilhari

Tags:    

Similar News