Introduction
On November 13, 2024 [2] [11], GEMA [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11], the German music rights organization [6] [11], initiated a landmark lawsuit against OpenAI [6] [7] [11], marking the first legal action by a European collecting society against an AI provider. The lawsuit addresses the unauthorized use of copyrighted song lyrics by OpenAI’s ChatGPT, highlighting broader concerns about AI’s impact on intellectual property rights and the need for fair compensation for creators.
Description
On November 13, 2024 [2] [11], GEMA [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11], the German music rights organization representing approximately 100,000 members, initiated a significant lawsuit against OpenAI [6] [7] [11], LLC and OpenAI Ireland Ltd in the Munich Regional Court [2]. This legal action marks the first of its kind by a collecting society in Europe against an AI provider [11]. GEMA alleges that OpenAI’s chatbot [5] [6] [11], ChatGPT [1] [4] [5] [6] [7] [8] [9] [10] [11], unlawfully reproduces copyrighted song lyrics from German authors without obtaining the necessary licenses or compensating the creators [4] [8]. The lawsuit contends that the companies have systematically infringed on copyrights by using protected texts, including lyrics from its members [5] [10], without proper authorization [1]. GEMA aims to demonstrate that OpenAI utilized its extensive repertoire for training purposes, as evidenced by the chatbot’s ability to reproduce original lyrics from simple prompts [10]. GEMA claims that ChatGPT was trained on lyrics from its members [4], further emphasizing the need for accountability in AI usage.
GEMA’s Managing Director [9], Tobias Holzmüller [9] [10], has stressed that the works of their members should not be treated as free resources for generative AI business models [1] [6] [9]. Unlike other internet services that compensate creators for their content [2], OpenAI allegedly engages in systematic copyright infringements by utilizing song lyrics as free resources, infringing on copyright and circumventing fair remuneration [2]. The case is supported by numerous prominent German music artists and their publishers [2], who contend that their lyrics have been exploited by the chatbot without any compensation [2]. To further protect its members, GEMA has opted out for all licensed works and declared an opt-out on their behalf, aiming to prevent large tech companies from solely benefiting from new technologies [8].
The lawsuit not only addresses the rights to song lyrics but also seeks to clarify various legal issues associated with AI usage [8], including unauthorized modifications generated by ChatGPT [9], which GEMA considers infringements of moral rights [9]. GEMA’s legal action reflects growing concerns among rightsholders globally about the impact of AI on the value of their works and the need for appropriate compensation [3]. Current German copyright law mandates that making protected works available to the public requires a license [7], and GEMA’s declaration of a reservation of rights makes the training of AI models on such material illegal under existing regulations.
In addition to pursuing this lawsuit, GEMA is actively engaging with AI providers to develop a licensing framework for the use of copyrighted content in AI training. GEMA has proposed a standard compensation model that includes a 30 percent revenue share for artists from AI-generated content [1], applicable to both initial training and subsequent use [11]. Although such a model has not yet been implemented in the market, GEMA has introduced an AI charter outlining ten principles for collaboration between AI companies and the music industry [4], advocating for responsible AI use [5]. This charter follows GEMA’s recent proposal for a generative AI licensing model [5], which aims to ensure fair participation for music creators when their works are used in AI training and content generation [7].
The legal context in Europe differs from that in the USA [7], where the “fair use” doctrine allows certain unlicensed uses [7], complicating the situation for European rights holders [7]. GEMA argues that European regulations require AI companies to obtain licenses for training on copyrighted works [11], contrasting with the US’s more lenient provisions [11]. Given that OpenAI operates in Germany [3], it is expected to comply with German copyright law [3], which offers clearer guidelines compared to the US system [3], where fair use can be ambiguous [3]. Additionally, European jurisdictions typically do not allow for the same level of damages as those available in the US [3]. GEMA utilizes AI for its internal operations and pays licensing fees [7], reinforcing its commitment to ensuring fair remuneration for human creators [7].
Furthermore, GEMA published a study in January 2024 examining the impact of generative AI on the music industry in Germany [7], indicating that generative AI services could jeopardize 27 percent of songwriters’ and lyricists’ income [11], potentially resulting in a loss of €2.7 billion by 2028 for its members and those of SACEM [11], the French counterpart [11]. The study reveals that a significant majority of creators believe AI poses a threat to their financial stability [11]. Other AI companies [1] [3] [4] [5] [6] [9] [11], such as Anthropic [1], Suno [5], and Udio [5], are also facing legal action for similar copyright issues [5], underscoring the broader implications of this legal landscape. Through this lawsuit [9], GEMA seeks to clarify the legal landscape surrounding AI and copyright [9], with a potential outcome involving the court referring questions to the European Court of Justice for further clarification [9]. Should GEMA succeed [9], OpenAI may face substantial financial liabilities and be mandated to remove all GEMA repertoire tokens from its models’ databases [9]. GEMA’s actions are intended to draw attention to OpenAI’s policies and the responsibilities of AI providers regarding royalties, emphasizing the potential threat to the livelihoods of creative professionals [4].
Conclusion
GEMA’s lawsuit against OpenAI underscores the growing tension between AI technology and intellectual property rights. The case highlights the urgent need for clear legal frameworks and fair compensation models to protect creators’ rights in the face of advancing AI technologies. The outcome of this legal action could set a precedent for how AI companies engage with copyrighted content, potentially influencing global standards and practices in the industry.
References
[1] https://www.faz.net/aktuell/wirtschaft/kuenstliche-intelligenz/urheberrecht-und-ki-gema-geht-gegen-open-ai-vor-110108268.html
[2] https://chatgptiseatingtheworld.com/2024/11/15/gema-sues-openai-in-germany-for-copyright-infringement/
[3] https://www.billboard.com/pro/gema-sues-openai-song-lyrics-copyright-law-europe/
[4] https://musically.com/2024/11/13/german-collecting-society-gema-sues-openai-over-song-lyrics/
[5] https://www.musicbusinessworldwide.com/openai-sued-by-gema-in-germany-for-unlicensed-use-of-song-lyrics/
[6] https://www.aibase.com/news/13227
[7] https://www.gema.de/en/news/ai-and-music/ai-lawsuit
[8] https://www.gema.de/de/aktuelles/ki-und-musik/ki-klage
[9] https://www.heise.de/en/news/GEMA-sues-OpenAI-Lyrics-built-on-theft-10032338.html
[10] https://www.yahoo.com/news/german-music-rights-body-takes-155721164.html
[11] https://the-decoder.com/german-music-rights-group-gema-sues-openai-over-unlicensed-song-lyrics-in-chatgpt/
