Copyright infringement risk for AI
Description
A model might generate content that is similar or identical to existing work protected by copyright or covered by open-source license agreement.
Why is copyright infringement a concern for foundation models?
Laws and regulations concerning the use of content that looks the same or closely similar to other copyrighted data are largely unsettled and can vary from country to country, providing challenges in determining and implementing compliance.
Parent topic: AI risk atlas
We provide examples covered by the press to help explain many of the foundation models' risks. Many of these events covered by the press are either still evolving or have been resolved, and referencing them can help the reader understand the potential risks and work toward mitigations. Highlighting these examples are for illustrative purposes only.