Landmark Court Rulings Clarify Copyright Boundaries for AI Training and Output
Recent federal court decisions establish that AI models cannot claim copyright protection for their outputs and that training on copyrighted material without authorization constitutes infringement. These rulings create new litigation opportunities for attorneys challenging unauthorized data use.
Over 25 copyright infringement lawsuits against AI companies are currently pending in federal court, testing copyright law on both the inputs used to train AI models and the outputs they generate. These disputes create new opportunities for plaintiffs attorneys to challenge the unauthorized use of copyrighted material in the rapidly evolving tech landscape.
A landmark February 2025 decision from the District of Delaware rules in favor of Thomson Reuters in its case against Ross Intelligence, marking the first major US ruling on AI training data. The court finds that Ross's use of Westlaw's legal headnotes harms the market for the original work and fails to meet the threshold of transformative use. Additionally, the US Court of Appeals for the DC Circuit upholds a lower court ruling in Thaler v. Perlmutter in March 2025, affirming that human authorship remains a strict prerequisite for copyright protection under US law.
The US Copyright Office addresses the copyrightability of AI-generated works in the second part of its Report on Copyright and Artificial Intelligence, published in January 2025. Drawing on over 10,000 public comments, the report provides critical conclusions regarding whether outputs created by generative AI qualify for protection. As these cases and federal guidelines set new precedents, plaintiffs attorneys play an essential role in shaping the future of copyright protections in the AI era.