Google Sued Over Using Copyrighted Books to Train Gemini AI

Google has been sued over allegations that it used copyrighted books without permission to train its Gemini artificial intelligence model, in the latest intellectual property case targeting a major technology company's AI development practices.

The lawsuit joins a rapidly expanding body of litigation in which authors, news organisations, and other content creators are challenging AI companies for harvesting their work as training data without seeking licences or paying compensation. Cases have been filed against OpenAI, Meta, and others, but a case targeting Gemini — Google's flagship AI product — carries particular weight given the company's scale and the model's widespread deployment.

At the heart of these cases is a fundamental question about copyright in the age of AI: does feeding a book, article, or piece of code into a machine-learning system constitute a reproduction or derivative use that requires the rightsholder's permission? The AI industry has generally argued that training on publicly available data constitutes fair use; rights holders strongly disagree.

For the broader AI ecosystem, the outcome of cases like this one could be transformative. If courts rule that training data must be licensed, the cost and complexity of building large language models would rise dramatically — potentially consolidating power among the largest players who can afford compliance. Indian courts and policymakers are watching these international proceedings closely as they shape India's own emerging AI regulatory framework.