As lawsuits involving the unauthorized use of copyrighted material to train generative artificial intelligence (GenAI) tools like ChatGPT continue to wend their way through the courts, copyright offices in the United States and the European Union are exploring the issue without waiting for the courts.

On the same day that the US Copyright Office released the pre-publication version of its report on GenAI training (as further discussed below), the European Union Intellectual Property Office (EUIPO) announced its own report on GenAI from a copyright perspective.

As the EUIPO explains,

This study explores the developments in GenAI from the perspective of EU copyright law. It is structured around three main components, (1) a technical, legal and economic analysis to further understand the functionality of GenAI and the implications of its development, as well as a detailed examination of copyright-related issues regarding the (2) use of content in GenAI services development and the (3) generation of content.

The main findings of the EUIPO report are:

• Access to high-quality content is central to the development of GenAI services. 
  
• No ‘one-size-fits all’ solution for copyright holders to protect their rights has emerged yet.
  
• Public authorities, such as national IP authorities and the EUIPO, may play a role by providing technical support (for copyright holders to reserve their rights, and for AI developers to effectively respect such reservations) as well as non-technical support (e.g., public awareness, forums for technical information sharing, providing information to the public on available solutions, trends and developments).

The main questions posed by the US Copyright Office’s report are:

Do any of the acts involved [in training GenAI] require the copyright owners’ consent or compensation? And to the extent they do, how can that feasibly be accomplished?

As the Copyright Office notes,

These issues are the subject of intense debate. Dozens of lawsuits are pending in the United States, focusing on the application of copyright’s fair use doctrine. Legislators around the world have proposed or enacted laws regarding the use of copyrighted works in AI training, whether to remove barriers or impose restrictions. 

The stakes are high, and the consequences are often described in existential terms. Some warn that requiring AI companies to license copyrighted works would throttle a transformative technology, because it is not practically possible to obtain licenses for the volume and diversity of content necessary to power cutting-edge systems. Others fear that unlicensed training will corrode the creative ecosystem, with artists’ entire bodies of works used against their will to produce content that competes with them in the marketplace. The public interest requires striking an effective balance, allowing technological innovation to flourish while maintaining a thriving creative community.

The Copyright Office requested public comment on these issues and received more than 10,000 responses.

The Copyright Office concluded that “creating and deploying a generative AI system using copyright-protected material involves multiple acts that, absent a license or other defense, may infringe one or more rights.”

For example:

• The process of training Generative AI models is generally preceded by massive amounts of web scraping that results in the creation of locally stored copies of millions or billions of copyrighted works. This copying during the acquisition and curation process implicates the copyright holder’s reproduction right.
  
• The GenAI training process also implicates the right of reproduction because the speed and scale of training requires developers to download the dataset and copy it to high-performance storage prior to training. Second, during training, works or substantial portions of works are temporarily reproduced as they are “shown” to the model in batches.
  

Also, noted the Copyright Office,

Generative AI models sometimes output material that replicates or closely resembles copyrighted works. Users have demonstrated that generative AI can produce near exact replicas of still images from movies, copyrightable characters, or text from news stories. Such outputs likely infringe the reproduction right and, to the extent they adapt the originals, the right to prepare derivative works. Some commenters noted that, depending on the content type and the audience, they may implicate the public display and public performance rights as well.

In the wake of this report, President Trump fired Register of Copyrights Shira Perlmutter, who had held her position since 2020, during the first Trump administration.

As CBS News reported, Perlmutter was appointed to the post by now-former Librarian of Congress Carla Hayden, who herself was fired by President Trump recently.

According to CBS, Democratic Rep. Joe Morelle of New York commented that there was “surely no coincidence he acted less than a day after [Perlmutter] refused to rubber-stamp Elon Musk’s efforts to mine troves of copyrighted works to train AI models.”

Just like the haiku above, we like to keep our posts short and sweet. Hopefully, you found this bite-sized information helpful. If you would like more information, please do not hesitate to contact us here: https://aeonlaw.com/contact-us/.