The Madras High Court has demanded full transparency in a growing scandal shaking Tamil Nadu's public library system. At the heart of the controversy are allegations of mass purchases of fake, and possibly AI-generated, books with public money.
The Heart of the Matter
The Madurai
Bench of the Madras High Court has taken a firm stance on what it calls a 'scandalous state of affairs' in the procurement of books for Tamil Nadu's 4,755 public libraries. On November 10, a division bench comprising Justices G.R. Swaminathan and M.D. Sumathi ordered the Directorate of Vigilance and Anti-Corruption (DVAC) to register an FIR into the alleged scam. The court directed the state government to identify and suspend the officials involved, freeze the bank accounts of suspect publishers, and remove the questionable books from circulation. The Director of Public Libraries has been ordered to submit a detailed affidavit by November 16, identifying the officials responsible for the procurement process from top to bottom.
What Sparked the Controversy?
The issue exploded into public view when renowned writer Perumal Murugan raised alarms about dubious books he found at the newly inaugurated Periyar Arivulagam library in Coimbatore. He and other researchers flagged hundreds of titles from previously unknown publishers like Agal Vilakku Pathippagam and Phoenix 2023. These books were riddled with factual errors, historical inaccuracies, and nonsensical content. Some bore the names of fictitious authors, while others falsely credited real people, like economist Jean Dreze, for work they never wrote. A cross-check of their International Standard Book Numbers (ISBNs) revealed that some were fake or belonged to entirely different publications.
The 'AI-Generated' Connection
The term 'AI-generated' has become central to the case because the poor quality and bizarre content of many books suggest they may have been mass-produced using artificial intelligence tools. These tools can generate plausible-sounding text, which is then packaged with fabricated author names and publisher details to be sold to unsuspecting institutions. While it has not been definitively proven that all the books were made by AI, the court has noted that such works would fall into a category of 'doubtful taste' and should not be procured with public funds. The scandal has exposed a critical vulnerability in public procurement systems, which can be exploited by entities selling low-quality, machine-generated content as legitimate literature.
A Breach of a Prior Order
This is not the first time the Madras High Court has intervened in the state's book procurement process. The current proceedings were initiated as a contempt petition, arguing that the Directorate of Public Libraries had wilfully disobeyed a comprehensive court order from June 2024. That earlier order, which stemmed from a case filed in 2019, had laid down a transparent policy for procuring books to prevent precisely this kind of fraud, including the repackaging of old books under new titles. The court expressed its dismay that despite these clear directives, the 'scam' had not only continued but escalated. The bench pointedly remarked, 'Actors change, actions do not', noting the problem has persisted across different political administrations.
What Happens Next?
With the November 16 deadline for the submission of purchase records and official affidavits, the case is at a critical juncture. The DVAC's investigation is now formally underway, and five senior officials from the Directorate of Public Libraries have already been suspended. Across the state, an estimated 70,674 suspect books have been identified and removed from library shelves. The court has made it clear that it expects full accountability, warning that stocking public libraries with bogus books infringes on the citizens' fundamental right to receive accurate information. The outcome of this case could set a major precedent for how public institutions across India verify and procure knowledge resources in an age of artificial intelligence.
















