- AI models like ChatGPT, Gemini, and Claude are trained on massive datasets including books, articles, research papers, and internet content, many of which are copyrighted works used without the authors’ knowledge or consent.
- According to legal experts, the question of whether using copyrighted books to train AI is legal remains without a definitive answer as current copyright laws have not kept pace with technology.
- In 2025, Judge William Alsup ordered Anthropic to pay $1.5 billion in compensation to a group of authors for using books illegally copied from pirate libraries to train its AI.
- However, the ruling determined that the act of training AI on copyrighted works itself is legal; Anthropic was penalized for the source of the data violating copyright, not for the training activity.
- The judge likened the process of AI learning from trillions of words to a writer reading many works to create new content, rather than direct copying.
- Some experts believe this ruling favors the AI industry because copyright law primarily regulates the act of copying rather than prohibiting reading, absorbing, or learning from a work.
- The U.S. Copyright Act was enacted in 1976, so courts must now interpret old regulations to handle new AI-related disputes.
- Many lawsuits revolve around the principle of “fair use,” in which the court considers the purpose of use, the degree of content transformation, the amount of the work used, and the impact on the market.
- In the case of Thomson Reuters v. Ross Intelligence, the court concluded that using data to build a directly competing AI product is not considered fair use.
- Another ruling in the case of Thaler v. Perlmutter determined that works created entirely by AI are not eligible for copyright, while raising new questions about how to define the degree of contribution between AI and humans.
📌 The legal framework for copyright and AI is still in its formative stages. While some recent rulings lean toward considering AI training as legal if it does not violate data collection methods, lawsuits involving fair use, direct competition, and authors’ rights continue to unfold. Current decisions will shape how the AI industry develops for years to come, but they have not yet created a unified legal precedent.
