Delhi High Court Rules OpenAI's AI Training on News Content is Fair Use
The Delhi High Court has ruled in favor of OpenAI, stating that training its AI models using news content from Asian News International (ANI) does not infringe copyright. The court determined that this practice falls under the "fair use" exception for research under India's Copyright Act. This ruling sets a major legal precedent in India, suggesting that blocking AI training on copyrighted data could hinder local AI development and harm the public interest. It highlights the growing global legal debate over balancing AI innovation with intellectual property rights. Judge Amit Bansal noted that ANI failed to prove ChatGPT's generated responses directly copied its copyrighted material. Furthermore, the court refused to issue an interim injunction, stating it would negatively impact India's own Large Language Model (LLM) development and users.
## BACKGROUND
Large Language Models (LLMs) are AI systems trained on vast amounts of text data to generate human-like language. Globally, AI companies like OpenAI face numerous lawsuits from publishers and content creators who argue that scraping copyrighted material for training without permission is illegal. Courts are currently navigating whether this training constitutes "fair use" or requires licensing.