Copyright Safety for Generative AI

Forthcoming in the Houston Law Review

Houston Law Review, Vol. 61, No. 2, 2023

53 Pages Posted: 4 May 2023 Last revised: 3 Jan 2024

See all articles by Matthew Sag

Matthew Sag

Emory University School of Law

Date Written: December 3, 2023

Abstract

Generative AI based on large language models such as ChatGPT, DALL·E-2, Midjourney, Stable Diffusion, JukeBox, and MusicLM can produce text, images, and music that are indistinguishable from human-authored works. The training data for these large language models consists predominantly of copyrighted works. This Article explores how generative AI fits within fair use rulings established in relation to previous generations of copy-reliant technology, including software reverse engineering, automated plagiarism detection systems, and the text data mining at the heart of the landmark HathiTrust and Google Books cases. Although there is no machine learning exception to the principle of non-expressive use, the largeness of likelihood models suggest that they are capable of memorizing and reconstituting works in the training data, something that is incompatible with non-expressive use.

At the moment, memorization is an edge case. For the most part, the link between the training data and the output of generative AI is attenuated by a process of decomposition, abstraction, and remix. Generally, pseudo-expression generated by large language models does not infringe copyright because these models “learn” latent features and associations within the training data, they do not memorize snippets of original expression from individual works. However, this Article identifies particular situations in the context of text-to-image models where memorization of the training data is more likely. The computer science literature suggests that memorization is more likely when: models are trained on many duplicates of the same work; images are associated with unique text descriptions; and the ratio of the size of the model to the training data is relatively large. This Article shows how these problems are accentuated in the context of copyrightable characters and proposes a set of guidelines for “Copyright Safety for Generative AI” to reduce the risk of copyright infringement.

Keywords: Copyright, Fair use, Generative AI, Copyright Safety, Foundation Models, Large language models

JEL Classification: K00

Suggested Citation

Sag, Matthew, Copyright Safety for Generative AI (December 3, 2023). Forthcoming in the Houston Law Review, Houston Law Review, Vol. 61, No. 2, 2023, Available at SSRN: https://ssrn.com/abstract=4438593 or http://dx.doi.org/10.2139/ssrn.4438593

Matthew Sag (Contact Author)

Emory University School of Law ( email )

1301 Clifton Road
Atlanta, GA 30322
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
2,716
Abstract Views
8,263
Rank
9,164
PlumX Metrics