Key facts
- Google sued for training Gemini AI on copyrighted books without permission
- Case raises core question of whether AI training data use constitutes copyright infringement
- Outcome could force AI industry to license training data
- Reported by MediaNama
Google is facing a lawsuit over allegations that it trained its Gemini AI model on copyrighted books without obtaining permission from authors or publishers. The case, reported by MediaNama, adds to a growing wave of litigation targeting the world's largest AI developers over how they source the data that powers their systems.
At the heart of the dispute is a question that courts in multiple jurisdictions are only beginning to grapple with: does ingesting copyrighted text to train an AI model constitute infringement, or does it fall under fair-use protections? Google, like most major AI developers, has argued that training on publicly available data is transformative and therefore legally defensible. Plaintiffs in such cases typically contend that their work is being commercially exploited without consent or compensation.
The stakes extend well beyond Google. OpenAI, Meta and other AI giants face similar lawsuits from authors, news organisations and visual artists. A ruling that goes against Google could compel the entire industry to negotiate licensing agreements before training future models, dramatically raising the cost of developing frontier AI systems.
For India, the implications are tangible. Indian publishers, regional-language authors and content creators whose work appears in digitised form online could potentially become parties to or beneficiaries of similar actions. It also raises questions about AI companies operating in India and whether Indian copyright law — which is broadly similar to international norms — provides comparable protections. The outcome of this US case is being watched closely by legal and tech communities worldwide.