This week, our trainee Patent Attorney, Euan Devlin, explores the developing legal turmoil facing OpenAI.
In a ground breaking legal development, OpenAI, the organization behind some of the most advanced language models, including GPT-3 & GPT-4, has found itself entangled in a copyright infringement lawsuit filed on 19 September 2023. The Authors Guild: a group of prominent authors including John Grisham, Elin Hilderbrand, George R.R. Martin, and more, have alleged that OpenAI unlawfully used their copyrighted works to train their language models, leading to the creation of derivative works without their consent.
This lawsuit has far-reaching implications for the field of artificial intelligence, copyright law, and creative industries in general and comes off a string of US copyright lawsuits filed against OpenAI by other writers. (1, 2)
Legal Risks of Generative AI
When AI models replicate or transform copyrighted works without proper authorization, it raises serious questions about intellectual property infringement.
Determining ownership of AI-generated content remains a complex issue. Is it the creator of the AI model, the person providing the prompt, or someone else?
AI platforms often train on vast datasets containing numerous unlicensed works. This raises concerns about whether AI users can input copyrighted or trademarked works into these platforms without permission.
Recent lawsuits, such as Andersen v. Stability AI et al. and Getty’s case against Stable Diffusion, underscore the legal challenges generative AI faces. These cases raise questions about what constitutes a “derivative work” under intellectual property laws and how the fair use doctrine applies.
The fair use doctrine allows limited use of copyrighted material without the owner’s consent for purposes like criticism, commentary, teaching, or research. It plays a crucial role in determining whether AI-generated content infringes on copyrights.
Google’s defense against a lawsuit, where it argued that its search engine’s text scraping constituted fair use, serves as a precedent in the tech-copyright intersection. It highlights the importance of the transformative use of copyrighted material.
The U.S. Supreme Court case involving the Andy Warhol Foundation and photographer Lynn Goldsmith could redefine copyright law by addressing what makes a work “transformative.” A decision against Warhol could impact AI-generated works.
The Authors Guild Lawsuit Summarised
The lawsuit, filed by a group of authors, alleges that OpenAI utilized their copyrighted literary works to train its language models without obtaining proper permissions. This raises several crucial questions about copyright infringement, fair use, and the boundaries of AI model training.
The plaintiffs contend that OpenAI’s language models have reproduced substantial portions of their copyrighted texts, thereby violating their exclusive rights as creators. They argue that this unauthorized copying constitutes copyright infringement. Furthermore, the Authors Guild allege that training a large language model on dataset that includes copyrighted material may be an infringement.
OpenAI may counterclaim that their usage falls under the doctrine of “fair use.” Fair use permits limited use of copyrighted material without the author’s consent for purposes such as criticism, commentary, news reporting, teaching, scholarship, or research. However, the line between fair use and infringement can be blurry, and courts often decide on a case-by-case basis.
The Author Guild also assert that OpenAI’s models have generated derivative works based on their novels. These unauthorized derivatives have the potential to devalue the original creations and harm the authors’ commercial interests.
Potential Implications of the Authors Guild lawsuit
The lawsuit highlights several noteworthy points that could affect writers, creators, and AI developers:
AI developers and companies need to be vigilant about the sources they use for training their models. The lawsuit underscores the importance of obtaining proper permissions or ensuring that their AI models are trained on public domain or licensed content.
The debate over fair use in AI-generated content is likely to intensify. As AI becomes more proficient at generating creative works, the courts will need to grapple with determining when AI-generated content constitutes fair use and when it infringes upon copyrights.
Authors and creators may have to be proactive in safeguarding their intellectual property. This lawsuit serves as a reminder that copyright protection remains crucial in an era where AI can replicate and manipulate text on a massive scale.
The lawsuit may prompt policymakers to consider regulations that address AI’s role in copyright infringement. Balancing technological advancements with the protection of intellectual property rights will be a complex task.
Conclusion
The OpenAI copyright lawsuit has brought to the forefront a pressing issue at the intersection of technology, creativity, and law. As AI models become increasingly sophisticated, addressing copyright concerns will be paramount. Writers and creators should remain informed about the evolving landscape of AI-generated content, while AI developers should exercise caution and adhere to copyright laws to avoid potential litigation.
The outcome of this lawsuit could set a precedent for future cases and shape the way AI is used to generate, modify, or replicate creative works. For now, it serves as a stark reminder that in the digital age, protecting intellectual property rights is as important as ever.
–Euan Devlin, Trainee Patent Attorney
At Scintilla, we help innovative companies get a grip on their intellectual property. Our unique commercial approach combines registration of patents and trade marks with strategic input so that IP can be a springboard for business growth. If you would like to discuss your IP needs, do contact us or book a free initial consultation!