The proliferation of Generative AI has raised significant issues with respect to intellectual property rights, particularly with regard to copyrights. Training Generative AI systems, like OpenAI’s ChatGPT, requires massive amounts of content or data. These systems have also integrated web search functionality, ensuring they have access to the most contemporaneous data (referred to as retrieval augmented generation or “RAG”). However, rather than providing blue links with snippets that direct the user to the creator’s website, these systems use third-party online content to provide the answer directly.  This is a profound change that is set to change the web as we know it. The primary issue facing Generative AI systems and content creators is Fair Use. Those who offer Generative AI systems claim using others’ content in this way is Fair Use and thus not subject to copyright infringement liability or damages.  Content creators claim this use substitutes for their copyrighted works, causing significant harm.

This presentation (hosted by LES USA-Canada's Washington DC Chapter) will provide perspective on the current issues before U.S. Courts related to Fair Use, copyright damages, and the developing market for AI training data.

Speaker: Eric Carnick, Managing Director, Ocean Tomo, a part of J.S. held