News
Attorneys
Practice Areas
Technologies
- Artificial Intelligence
Anthropic's $1.5B Lesson: Fair Use Has Limits in the Age of AI
Michelle Ciotola and David Kincaid explore the recent settlement in Anthropic, which provides insights into the lawful use of copyrighted works for AI training and the risks associated with misuse in their article, "Anthropic's $1.5B Lesson: Fair Use Has Limits in the Age of AI," in The Trademark Lawyer magazine.
David and Michelle are frequent co-authors on AI and IP, bringing a cross-practice perspective to the issues that arise from the intersection of AI technology and intellectual property rights. David is a patent attorney and co-chair of the firm’s Artificial Intelligence Practice Group, and Michelle is Chair of the firm's Trademark and Copyright Practice Group.
From their article:
"Fast forward to August 2025: the parties reached a USD 1.5 billion settlement, ending the feud. What do the case and settlement teach us? The source of the works used in an LLM’s training is pivotal. Use of copyrighted works to train AI will likely be considered fair use if the works are lawfully acquired. But the use of pirated copies in the development of datasets is problematic and likely to violate US Copyright Law. It will be prudent to carefully track how datasets are acquired and avoid the use of shadow libraries."
