03.07.2024
news
What’s it all about?
Last weeks events mark a critical point in the ongoing conflict between the traditional music industry and emerging AI technologies and will most certainly influence how we see not only music, but any form of art that is created by artificial intelligence. Since the topic is somewhat complex and not immediately understandable for everyone, I have compiled a brief summary of the recent developments, gathered some background information, and written a small conclusion on what the lawsuit means for the use of AI-generated music in television, film, and especially advertising.
Background
Udio and Suno, both AI-driven music platforms, are under fire for allegedly using copyrighted music to train their AI models without proper authorization. Udio is particularly known as the company behind the AI-generated track "BBL Drizzy," while Suno has partnered with Microsoft to integrate its technology into the Copilot suite, allowing users to generate music.
The Allegations
The Recording Industry Association of America (RIAA) claims that both companies trained their AI models with a vast amount of copyrighted music without obtaining the necessary licenses. The RIAA states that despite demands for transparency regarding the training data, Udio and Suno have claimed confidentiality and insist that their tools are transformative and designed to create entirely new content rather than replicate existing works.
Evidence of Infringement
The RIAA's lawsuit includes examples of AI-generated tracks that are nearly identical to well-known songs such as Chuck Berry's "Johnny B. Goode" and Mariah Carey's "All I Want For Christmas Is You." Here are a few examples, the associated prompts, and the filed lawsuit: https://www.404media.co/listen-to-the-ai-generated-ripoff-songs-that-got-udio-and-suno-sued/ The RIAA argues that this level of similarity indicates direct copying rather than fair use or transformative creation.
The Legal and Ethical Debate
This case highlights a broader debate about the legality and ethics of using copyrighted material to train AI models. On one hand, AI companies argue that their practices fall under fair use because the transformation in creating new outputs justifies their actions. On the other hand, the music industry argues that these practices undermine the economic value of original works and violate the rights of creators.
The Unique Position of the Music Industry
Unlike other creative sectors, the music industry has a long history of aggressively defending its copyrights. This legal strictness is partly due to the well-established systems for monetizing music and compensating artists, which are now being challenged by AI technologies. The industry's organized approach to litigation, combined with the public's emotional connection to music, provides a strong foundation for these legal battles.
Strong Market Arguments
One of the central foundations of copyright law is its role in maintaining the economic value of creative works. The unauthorized use of copyrighted music to train AI models devalues the market for original music. This is not just a matter of lost revenue; it is a fundamental threat to the livelihoods of artists and the integrity of the music industry. The labels argue that AI-generated music trained on copyrighted works without permission directly competes with legitimate music, dilutes the market, and reduces potential revenue for original artists.
Further Implications
The outcomes of these lawsuits could set important precedents for how AI-generated content will be regulated and monetized in the future. A ruling against Udio and Suno could lead to stricter controls and higher costs for AI companies wishing to use copyrighted material, possibly stifling innovation. Conversely, a ruling in favor of the AI companies could open the door to broader use of copyrighted content in AI training and change the landscape of the creative industries. Based on recent rulings (e.g., the Marvin Gaye Estate vs. Pharrell Williams and Robin Thicke / “Blurred Lines” case), a decision in favor of the AI companies seems unlikely.
Significance for Synchronizing AI-Generated Music
For now, AI-generated music is in a legal limbo. While such tracks may find some streams, they are unlikely to be synchronized in film and television. This is especially true in connection with brands—no responsible advertiser will use a song with unclear rights and the potential risk of hidden copyright infringements to sell to a client. Additionally, the outputs of these AI models cannot be copyrighted—therefore, they cannot be used exclusively since theoretically anyone can create the same track with a corresponding prompt. Consequently, a viable business remains for composers and musicians who create original works for TV, film, and advertising, ensuring that human creativity and artistic integrity continue to have a place in the industry.
Sensible Use of AI in Music Production
The general legal uncertainty does not mean that AI cannot still be a very useful tool for creating new music. For instance, at BLUT, we have been using programs like ChatGPT, Cyanite, Emvoice, or LALALAI in our workflows for some time. AI can already make an important contribution at every stage of production to better and more cost-effectively capture ideas for clients: from writing and re-writing lyrics to stem separation of unlicensed moods to adapt them onto a film, from searching for comparable titles in a library (similarity search) to voice generation when layouting vocal tracks.
Back to the future
The music industry is by no means regressive or technophobic—the RIAA has welcomed every new technology so far, as long as the associated use of their copyrights could be monetized. For the Association, streaming is THE success story of the 21st century. And even if the fight of the record companies against companies like Udio or Suno is primarily about securing financial compensation, the preservation of the value and recognition of human artistry in an increasingly automated world is a very positive side effect.
03.07.2024
news
What’s it all about?
Last weeks events mark a critical point in the ongoing conflict between the traditional music industry and emerging AI technologies and will most certainly influence how we see not only music, but any form of art that is created by artificial intelligence. Since the topic is somewhat complex and not immediately understandable for everyone, I have compiled a brief summary of the recent developments, gathered some background information, and written a small conclusion on what the lawsuit means for the use of AI-generated music in television, film, and especially advertising.
Background
Udio and Suno, both AI-driven music platforms, are under fire for allegedly using copyrighted music to train their AI models without proper authorization. Udio is particularly known as the company behind the AI-generated track "BBL Drizzy," while Suno has partnered with Microsoft to integrate its technology into the Copilot suite, allowing users to generate music.
The Allegations
The Recording Industry Association of America (RIAA) claims that both companies trained their AI models with a vast amount of copyrighted music without obtaining the necessary licenses. The RIAA states that despite demands for transparency regarding the training data, Udio and Suno have claimed confidentiality and insist that their tools are transformative and designed to create entirely new content rather than replicate existing works.
Evidence of Infringement
The RIAA's lawsuit includes examples of AI-generated tracks that are nearly identical to well-known songs such as Chuck Berry's "Johnny B. Goode" and Mariah Carey's "All I Want For Christmas Is You." Here are a few examples, the associated prompts, and the filed lawsuit: https://www.404media.co/listen-to-the-ai-generated-ripoff-songs-that-got-udio-and-suno-sued/ The RIAA argues that this level of similarity indicates direct copying rather than fair use or transformative creation.
The Legal and Ethical Debate
This case highlights a broader debate about the legality and ethics of using copyrighted material to train AI models. On one hand, AI companies argue that their practices fall under fair use because the transformation in creating new outputs justifies their actions. On the other hand, the music industry argues that these practices undermine the economic value of original works and violate the rights of creators.
The Unique Position of the Music Industry
Unlike other creative sectors, the music industry has a long history of aggressively defending its copyrights. This legal strictness is partly due to the well-established systems for monetizing music and compensating artists, which are now being challenged by AI technologies. The industry's organized approach to litigation, combined with the public's emotional connection to music, provides a strong foundation for these legal battles.
Strong Market Arguments
One of the central foundations of copyright law is its role in maintaining the economic value of creative works. The unauthorized use of copyrighted music to train AI models devalues the market for original music. This is not just a matter of lost revenue; it is a fundamental threat to the livelihoods of artists and the integrity of the music industry. The labels argue that AI-generated music trained on copyrighted works without permission directly competes with legitimate music, dilutes the market, and reduces potential revenue for original artists.
Further Implications
The outcomes of these lawsuits could set important precedents for how AI-generated content will be regulated and monetized in the future. A ruling against Udio and Suno could lead to stricter controls and higher costs for AI companies wishing to use copyrighted material, possibly stifling innovation. Conversely, a ruling in favor of the AI companies could open the door to broader use of copyrighted content in AI training and change the landscape of the creative industries. Based on recent rulings (e.g., the Marvin Gaye Estate vs. Pharrell Williams and Robin Thicke / “Blurred Lines” case), a decision in favor of the AI companies seems unlikely.
Significance for Synchronizing AI-Generated Music
For now, AI-generated music is in a legal limbo. While such tracks may find some streams, they are unlikely to be synchronized in film and television. This is especially true in connection with brands—no responsible advertiser will use a song with unclear rights and the potential risk of hidden copyright infringements to sell to a client. Additionally, the outputs of these AI models cannot be copyrighted—therefore, they cannot be used exclusively since theoretically anyone can create the same track with a corresponding prompt. Consequently, a viable business remains for composers and musicians who create original works for TV, film, and advertising, ensuring that human creativity and artistic integrity continue to have a place in the industry.
Sensible Use of AI in Music Production
The general legal uncertainty does not mean that AI cannot still be a very useful tool for creating new music. For instance, at BLUT, we have been using programs like ChatGPT, Cyanite, Emvoice, or LALALAI in our workflows for some time. AI can already make an important contribution at every stage of production to better and more cost-effectively capture ideas for clients: from writing and re-writing lyrics to stem separation of unlicensed moods to adapt them onto a film, from searching for comparable titles in a library (similarity search) to voice generation when layouting vocal tracks.
Back to the future
The music industry is by no means regressive or technophobic—the RIAA has welcomed every new technology so far, as long as the associated use of their copyrights could be monetized. For the Association, streaming is THE success story of the 21st century. And even if the fight of the record companies against companies like Udio or Suno is primarily about securing financial compensation, the preservation of the value and recognition of human artistry in an increasingly automated world is a very positive side effect.