The US Copyright Office is on the verge of a significant shift in policy that could redefine how generative AI interacts with copyrighted works. With a public comment period now open until October 18, the Office is seeking to gather insights that will guide its ongoing study of artificial intelligence tools such as Midjourney, OpenAI’s ChatGPT, and Google Bard. This move signifies a step closer to proposing new rules or regulatory action in response to the growing public discourse surrounding the implications of these systems on the future of creative industries.
The intensifying debate finds its roots in the use of copyrighted works to train AI models, the level of transparency and disclosure needed in using these works, and the legal status of AI-generated outputs. Authors, visual artists, and even source code developers have already instigated lawsuits against tech giants like OpenAI, Microsoft, and Meta for using their original work without consent, stoking fears of unfair competition. These concerns, coupled with the broader implications of AI in the creative industry, have prompted the Copyright Office to consider whether new rules or regulations are necessary. The public comment period, a typical precursor to the proposal of a final rule, represents the culmination of earlier "listening sessions" with stakeholders and the start of a potentially transformative chapter in AI copyright regulation.
US Copyright Office Seeks Public Opinion on Generative AI Rules
The US Copyright Office is moving towards establishing new rules and guidelines concerning the use of generative artificial intelligence (AI) and its impact on the works of authors and creators. The Office has opened a public consultation period until October 18 to gain further insight into the ongoing study of AI tools like Midjourney, OpenAI’s ChatGPT, and Google Bard.
A Call for Public Opinion on Generative AI Regulation
The Office is considering potential regulatory action or new federal rules in response to the ongoing public conversation about the future implications of these systems for creative industries. They are seeking opinions on whether new rules or regulations for generative AI are necessary, as well as issues related to the use of copyrighted works to train AI models, the appropriate levels of transparency and disclosure with respect to the use of copyrighted works, and the legal status of AI-generated outputs.
Legal Actions Against Tech Giants
Several authors, visual artists, and source code developers have taken legal action against major tech companies such as OpenAI, Microsoft, and Meta, accusing them of using their original work without permission to train AI systems that could potentially compete against them. This public comment period typically precedes the proposal and adoption of a final rule. Earlier this year, the Office conducted "listening sessions" with stakeholders, including representatives from Microsoft, VC firm Andreessen Horowitz, and The Authors Guild.
The Impact of Web Scraping on Content Creators
Generative AI tools like ChatGPT, Bard, and Meta’s Llama 2 have been developed using massive volumes of information collected by web crawlers. These crawlers gather as much online data as possible, including millions of copyrighted works. The tech companies behind these generative AI tools use the collected data to train their models without compensating the original content creators.
As more online businesses become aware of the extent of web scraping for the benefit of generative AI, companies like Amazon, Airbnb, Tumblr, and news outlets like The New York Times and CNN are increasingly blocking web crawlers using a tool known as robots.txt. Currently, there is almost no way to prevent content from being crawled and used to create an LLM, and existing copyright law does not address this issue.
Future Considerations for AI and Copyright Law
The US Copyright Office has already made several decisions on AI-related content issues. In one instance, it declined to register a copyright for an AI-generated artwork. In another case involving a book with AI-generated imagery by a human author, the Office only granted copyright for the author’s text.
The Office is now considering additional issues related to copyright and AI, including training data, the eligibility of AI-generated material for copyright, liability for infringement by AI systems, and how to handle AI outputs that "imitate the identity or style of human artists."
This last issue is particularly pertinent due to ongoing strikes by Hollywood writers and actors who are partly protesting against the potential use of generative AI in place of human writers and performers.
The emergence of generative AI has created a complex landscape for copyright law and creative industries. The ongoing public comment period by the US Copyright Office signifies a crucial step towards understanding and addressing these complexities. It also underscores the need for clear regulations that balance the interests of tech companies, content creators, and the broader public. As AI continues to evolve, it’s essential to continue these conversations to ensure fair use and recognition of original works in the digital age.