Copyright and Artificial Intelligence: An Overview of How They Intersect

9:55 AM - 10:10 AM

Copyright is present throughout AI lifecycles, including the use of copyrighted works for machine training and the creation of new works. This presentation will focus on the inputs rather than the outputs. Large Language Models and other AI technologies rely on millions of books, scholarly journals, and other curated publications. Using and re-using these works responsibly and legally is a foundational part of the discussion about Generative AI. A few basic copyright concepts will be introduced to provide some context, followed by a brief review of key court cases that have been initiated lately and of the regulatory framework in place or under discussion in different countries. Finally, reference will be made to how voluntary non-exclusive collective licensing of copyright can be used alongside direct licensing as an efficient way to enable authorized uses of large collections of copyrighted materials for AI purposes and to reduce the risk of infringement.