Introduction
The ongoing legal battles between content publishers and AI companies over copyright infringement are reshaping the landscape of intellectual property law. These cases [1] [2] [3] [4] [6], which involve significant questions about the use of copyrighted materials for training AI models, have the potential to fundamentally alter the information ecosystem and the AI industry [6].
Description
On August 22, 2024 [1], Judge Bibas granted a continuance in the Thomson Reuters v [1]. ROSS Intelligence case [1] [2] [4], delaying the first trial in this significant AI copyright lawsuit [1]. Thomson Reuters initiated the lawsuit in May 2020, accusing ROSS Intelligence [1], a legal AI startup [2] [6], of violating US copyright law by unauthorized reproduction of materials from its Westlaw platform [2] [4] [6]. Initially gaining little attention, this case has since emerged as a significant precursor to ongoing legal battles between content publishers and AI companies [6], raising critical questions about the use of copyrighted materials for training AI models. The implications of these lawsuits could fundamentally alter the information ecosystem and the AI industry [6], affecting a wide range of stakeholders across the internet [6]. The ongoing litigation has already led to the closure of ROSS due to mounting legal costs, and the case is poised to produce the first final judgment on fair use in the context of AI model training [1]. A decision on renewed motions for summary judgment is expected in January 2025 [1], with the trial now scheduled for May 12, 2025 [1].
Pivotal developments in copyright lawsuits are anticipated in 2024 [3], potentially influencing the future of artificial intelligence businesses [3]. Various lawsuits have been filed by authors [3], news outlets [3], visual artists [2] [3] [4], musicians [3], and other copyright holders against technology companies [3], including major generative AI firms like OpenAI, Meta [1] [2] [3] [4] [5], and Anthropic [3]. These lawsuits allege unauthorized use of copyrighted works to train AI content generators [3], raising the central legal issue of whether the defendants’ actions constitute “fair use.” Tech companies assert that their AI systems utilize copyrighted material in a transformative manner [3], while copyright holders contend that these companies unlawfully replicate their works [3], undermining their livelihoods [3]. Some content owners have begun to license their material to tech companies [3], while others [3] [5], such as major record labels and prominent authors [3], continue to pursue legal action [3].
In a related development [1], OpenAI filed a petition on December 6, 2024, to consolidate eight ongoing copyright lawsuits under 28 US Code Section 1407 [5], seeking to centralize oversight of these cases to enhance efficiency and consistency in pretrial proceedings. The lawsuits involve various plaintiffs, including authors [2] [4] [5], YouTube creators [5], and news organizations [5], alleging that OpenAI used their copyrighted works without compensation to train its large language models (LLMs), resulting in outputs that summarize or quote those works [5]. OpenAI argues that its activities are protected by the doctrine of fair use [5], although no substantive court decisions have yet clarified whether training LLMs on copyrighted material constitutes infringement [5]. Preliminary rulings on motions to dismiss have indicated that judges find the plaintiffs’ claims worthy of trial and discovery [5].
The outcome of these cases may hinge on varying interpretations of fair use across different jurisdictions [3], with potential for conflicting rulings and multiple appeals [3]. The fair use issue remains a focal point in these lawsuits [1], with courts showing varying responses to similar claims in previous cases. In a notable development [1], David Boies and his firm were appointed as interim lead class counsel in a case involving Meta [1], following a rebuke that led to a change in legal representation [1]. Boies’s team deposed Meta CEO Mark Zuckerberg [1], claiming to have obtained damaging testimony that supports filing a Third Amended Complaint [1], which includes a DMCA CMI claim and a California state law claim [1].
The interpretation of Section 1202 of the Digital Millennium Copyright Act (DMCA) regarding the removal of copyright management information (CMI) is also critical [1]. Unlike copyright infringement claims [1] [5], DMCA CMI claims do not require copyright registration [1]. However, district courts have largely dismissed these claims [1], with various judges citing reasons such as lack of evidence for the removal of CMI or lack of standing [1]. In the Intercept v [1]. OpenAI case [1] [2] [4] [5], some DMCA claims were dismissed [1], but a claim regarding intentional removal of CMI was allowed to proceed [1]. The implications of the Ninth Circuit’s decisions in these matters are significant [1], shaping the landscape of copyright law as it pertains to AI technologies. Other notable cases [2] [4], including The New York Times’ lawsuit against OpenAI and Microsoft [2] [4], are currently in contentious discovery phases [2] [4], further highlighting the ongoing legal challenges faced by AI companies.
Conclusion
The outcomes of these legal proceedings will have far-reaching implications for the AI industry and the broader information ecosystem. As courts grapple with the complexities of fair use and copyright law in the context of AI, the decisions made in these cases will likely set precedents that influence future interactions between technology companies and content creators. The evolving legal landscape will require stakeholders to navigate new challenges and opportunities as they adapt to the changing dynamics of intellectual property rights in the digital age.
References
[1] https://chatgptiseatingtheworld.com/2024/12/31/the-top-5-stories-in-u-s-copyright-lawsuits-v-ai-in-2024/
[2] https://www.wired.com/story/ai-copyright-case-tracker/
[3] https://lnginnorthernbc.ca/2024/12/27/tech-companies-will-face-ai-copyright-challenges-in-2025/
[4] https://bestofai.com/article/every-ai-copyright-lawsuit-in-the-us-visualized
[5] https://roninlegal.in/openais-seeks-to-consolidate-copyright-lawsuits-a-closer-look/
[6] https://bioethics.com/archives/95192
