Major American Music Labels Sue Generative AI Music Platforms in First Case of Its Kind Over AI Audio
What You Need to Know
Key takeaway #1
A group of major music labels has banded together to sue the leading generative AI music creation websites, alleging that the websites unlawfully trained their AI on the labels’ copyrighted sound recordings.
Key takeaway #2
The lawsuit is the first of its kind against a music/sound based generative AI program, as opposed to pending lawsuits targeting text and image generation.
Client Alert | 4 min read | 06.25.24
Universal Music Group, Sony Music, and Warner Music Group., represented by the Recording Industry Association of America (RIAA), have sued online music AI generators, Suno AI (“Suno”) and Udio AI (“Udio”), for alleged copyright infringement, accusing them of replicating their artists’ music using AI technology. The Suno complaint is filed in the U.S. District Court for the District of Massachusetts, and the Udio complaint is filed in the U.S. District Court for the Southern District of New York. The lawsuits also target Alphabet Inc., Google's parent company. The RIAA is asking for damages amounting to up to $150,000 per infringing song, which could amount to hundreds of millions of dollars.
Suno and Udio are text-to-music AI generators, which allow users to create full-length songs with instrumentation and lyrics based on what the user inputs into the program. For example, music and tech publications have tested Suno by inputting short text phrases, like TechRadar asking to “generate a song celebrating gadgets and technology in the genre of electronic chillwave,” and having Suno produce songs like “Future Frequencies” or Rolling Stone requesting a “delta blues track,” and producing “Soul of the Machine.” AI generated songs, such as these, among others, have shaken the music industry. Music generation programs use both “diffusion” and “transformer” models to publish an AI output, i.e., both combining and creating musical elements on the back end of the program in order to generate “novel” outputs.
This will be an important case to watch for both copyright and AI practitioners in the United States, as this is the first lawsuit that tackles the generation of audio, combined with the extensive combination of a sound recording—implicating genres, styles, rhythms, specific instruments, lyrics, and sound-alike vocals. The case may implicate the bounds of fair use of copyright in relation to AI generation, and may force courts to grapple with what constitutes “musical elements” of rhythm and melody in relation to current songs and artists.
On its website, Suno acknowledges the current issues with AI generation protection: “the availability and scope of copyright protection for content generated (in whole or in part) using artificial intelligence is a complex and dynamic area of law, which is rapidly evolving and varies among countries." Similarly, Suno prohibits uploading lyrics from copyrighted songs as the text input for new generations: “[You cannot upload lyrics to existing songs] unless you are the copyright owner of those existing lyrics or you get explicit permission from the copyright owner to upload their existing lyrics into Suno.” Udio’s website includes similar caveats on AI generation protection, explaining that “[t]he law regarding copyright protection over AI-generated works is complex and constantly changing, and the law may vary by jurisdiction.” According to their policy, they “explicitly forbid the use of copyrighted material and any other third-party intellectual property to generate content using Udio […].”
In the Suno complaint, RIAA acknowledges that there is both “promise and peril with AI,” but contends that Suno has overstepped the permissible bounds of copyright while training its AI engine: “if developed irresponsibly, without regard for fundamental copyright protections, those same tools threaten enduring and irreparable harm to recording artists, record labels, and the music industry, inevitably reducing the quality of new music available to consumers and diminishing our shared culture.” The complaint continues to allege copyright infringement by claiming that Suno copied “sound recordings en masse and ingested them into [their] AI model[s]. [These] product[s] can only work the way [they do] by copying vast quantities of sound recordings from artists across multiple genres, styles, and eras.” Likewise, the Udio complaint mirrors language on the promise and peril of AI, and emphasizes that “Udio’s product can only work the way it does by copying vast quantities of sound recordings from artists across every genre, style, and era.”
In a press release by RIAA, the company states, “these lawsuits are necessary to reinforce the most basic rules of the road for the responsible, ethical, and lawful development of generative AI systems and to bring Suno’s and Udio’s blatant infringement to an end.” As such, the cases seek (1) declarations that defendants infringed on plaintiff’s copyrighted sound recordings; (2) injunctions barring defendants from future infringement; and (3) damages for the infringement that has already occurred.
If the courts side with the music companies, companies such as Suno and Udio may need to alter their business models. This could also prompt changes in copyright laws, which may create tougher guidelines for the use of copyrighted content on AI platforms.
As cases move forward, Crowell & Moring, LLP will continue to follow AI in the courts, advising clients on copyright compliance for AI.
Contacts
Insights
Client Alert | 8 min read | 12.20.24
End of Year Regulations on Interoperability
Federal policy efforts to advance health data exchange and interoperability are continuing to change rapidly. The latest changes are the publication of two final rules by the Assistant Secretary for Technology Policy/Office of the National Coordinator for Health Information Technology (ASTP/ONC) finalizing parts of the of the Health Data, Technology, and Interoperability (HTI-2) Proposed Rule. These rules adopt requirements regarding the Trusted Exchange Framework and Common Agreement (TEFCA) (HTI-2 Part 1), and create a new Information Blocking exception under Protecting Care Access (HTI-2 Part 2), on December 16th and 17th, respectively.
Client Alert | 4 min read | 12.19.24
Client Alert | 4 min read | 12.19.24
Key Changes to the State Attorneys General – 2024 to 2025 Transition
Client Alert | 4 min read | 12.19.24
New EU Directive Impacting Digital Platforms and Individuals Working for Them