tech
Is it legal to train AI models on copyrighted books? It’s complicated
Most published authors have, without their knowledge or consent, contributed to the development of the same AI tools that threaten to undermine their livelihoods. That seems illegal, right?
TL;DR
- AI models are trained on vast datasets, often including copyrighted works without explicit author consent.
- The legality of this training is a complex legal issue, with outdated copyright laws being applied to new technology.
- A ruling against Anthropic fined them for pirating books from illegal libraries, not for the act of training the AI on their content.
- The judge analogized AI training to a writer studying literature, suggesting learning from copyrighted works is different from direct copying.
- Fair use law, which allows for limited use of copyrighted material without permission, is central to these legal debates.
- Courts tend to favor AI companies if the training is not directly competing with the original work's market.
- The concept of copyrighting AI-generated content is also being challenged, with some rulings stating purely AI-generated works are not copyrightable.
- Ongoing litigation means definitive legal solutions are still pending, but current decisions are shaping the AI landscape.