MINING WITHOUT A MAP: TEXT-AND-DATA-MINING, AI TRAINING, AND THE LIMITS OF FAIR DEALING UNDER INDIAN COPYRIGHT LAW AFTER ANI V. OPEN AI

Abstract

The development of large language models relies extensively on text obtained from the open internet, much of which is copyright-protected and may be incorporated into training datasets without authors’ consent or remuneration. The resulting question is whether such text-and-data mining (TDM) constitutes copyright infringement or falls within an existing statutory exception. India first received substantial judicial treatment of this issue in Asian News International v OpenAI, decided by the Delhi High Court on 24 July 2026 at the interim stage. The Court held, prima facie, that the storage and use of ANI’s literary works for training the large language models underlying ChatGPT satisfied the purpose and fairness requirements of section 52(1)(a). This paper critically examines that reasoning and argues that the decision demonstrates both the adaptability and limits of India’s fair-dealing framework. Although section 52 does not establish the open-ended fair-use doctrine found in the United States, the Court broadly interpreted “research”. It developed a distinct fairness inquiry to accommodate AI training. The paper contends that this approach provides an immediate judicial response but leaves unresolved questions concerning the scale of AI training, market substitution, lawful access, and the institutional limits of judicial adaptation. A comparison with the European Union, Japan, the United Kingdom and the United States demonstrates materially different approaches to reconciling TDM with copyright. Against this background, the paper evaluates the Department for Promotion of Industry and Internal Trade (DPIIT) licensing model and argues that a remunerated legislative framework may offer a more durable solution than relying exclusively on case-by-case judicial development of section 52.

Keywords: Text-and-Data Mining (TDM), Fair Dealing, Section 52, Artificial Intelligence Training, Statutory Licensing.