A court in Munich has determined that OpenAI’s ChatGPT breached German copyright laws by utilizing popular songs from renowned artists to train its language model, which advocates for the creative industry have labeled a pivotal ruling for Europe.
The Munich District Court supported the German music copyright association GEMA, stating that ChatGPT gathered protected lyrics from well-known musicians to “learn” them.
GEMA, an organization that oversees the rights of composers, lyricists, and music publishers with around 100,000 members, initiated legal action against OpenAI in November 2024.
This case was perceived as a significant test for Europe in its efforts to prevent AI from harvesting creative works. OpenAI has the option to appeal the verdict.
ChatGPT lets users pose inquiries and issue commands to a chatbot, which replies with text that mimics human language patterns. The foundational model of ChatGPT is trained on widely accessible data.
The lawsuit focused on nine of the most iconic German hits from recent decades, which ChatGPT employed to refine its language skills.
This included Herbert Groenemeyer’s 1984 synthpop hit manners (male), and Helen Fischer’s Atemlos Durchi die Nacht (Breathless Through the Night), which became the unofficial anthem for the German team during the 2014 World Cup.
The judge ruled that OpenAI must pay undisclosed damages for unauthorized use of copyrighted materials.
Kai Welp, GEMA’s general counsel, mentioned that GEMA is now looking to negotiate with OpenAI about compensating rights holders.
The San Francisco-based company, co-founded by Sam Altman and Elon Musk, argued that its language learning model utilizes the entire training set rather than retaining or copying specific songs, as stated by the Munich court.
OpenAI contended that since the outputs are created in response to user prompts, the users bear legal responsibility, an argument the court dismissed.
GEMA celebrated the ruling as “Europe’s first groundbreaking AI decision,” indicating that it might have ramifications for other creative works.
Tobias Holzmuller, the company’s CEO, remarked that the verdict demonstrates that “the internet is not a self-service store, and human creative output is not a free template.”
“Today, we have established a precedent to safeguard and clarify the rights of authors. Even AI tool operators like ChatGPT are required to comply with copyright laws. We have successfully defended the livelihood of music creators today.”
The Berlin law firm Laue, representing GEMA, stated that the court’s ruling “creates a significant precedent for the protection of creative works and conveys a clear message to the global tech industry,” while providing “legal certainty for creators, music publishers, and platforms across Europe.”
The ruling is expected to have ramifications extending beyond Germany as a legal precedent.
The German Journalists Association also praised the decision as a “historic triumph for copyright law.”
OpenAI responded that it would contemplate an appeal. “We disagree with the ruling and are evaluating our next actions.” The statement continued, “This ruling pertains to a limited set of lyrics and does not affect the millions of users, companies, and developers in Germany who utilize our technology every day.”
Furthermore, “We respect the rights of creators and content owners and are engaged in constructive discussions with various organizations globally that can also take advantage of this technology.”
OpenAI is currently facing lawsuits in the U.S. from authors and media organizations alleging that ChatGPT was trained on their copyrighted materials without consent.
Source: www.theguardian.com
