<\/p>\n
A group of independent musicians and songwriters sued Google<\/a> on Friday (March 6), accusing the company of training its new Lyria 3 music-generation model on copyrighted recordings pulled from YouTube<\/a> without permission or payment.<\/p>\n The lawsuit came over two weeks after Google<\/b> announced<\/a> the launch of Lyria 3 <\/b>on February 18, describing it as the \u2018most advanced\u2019 generative AI music model yet. The model operates within Google\u2019s Gemini<\/b> chatbot app, allowing users to create 30-second tracks from text prompts or images.<\/p>\n Lyria 3 was developed by\u00a0Google DeepMind<\/b>, the tech giant\u2019s AI research division, and is the latest iteration of its generative music technology.<\/p>\n In their 118-page lawsuit filed in the US District Court for the Northern District of Illinois (read<\/b> here<\/b><\/a>), the group of independent artists argued that \u201cGoogle copied millions of copyrighted sound recordings, musical compositions, and lyric, including at least 44 million clips and 280,000 hours of music,\u201d and used these to train Lyria 3.<\/p>\n The all-indie roster of plaintiffs includes New York-based singer and songwriter Sam<\/b> Kogon<\/b>, along with plaintiffs who also sued<\/a> AI music generators Udio and Suno in October: David<\/b> Woulard<\/b>, indie R&B group Attack the Sound<\/b>, father-and-son folk rock duo Stan and James Burjek<\/b>, Chicago-based band Directrix<\/b> members Berk Ergoz, Hamza Jilani, Maatkara Wilson <\/b>and Arjun Singh<\/b>, as well as Los Angeles-based composer Magnus<\/b> Fiennes<\/b>, and Atlanta-based songwriter and producer Michael<\/b> Mell<\/b>.<\/p>\n The latest lawsuit names Google as the sole defendant. The plaintiffs structured their lawsuit as a proposed class action.<\/p>\n \u201cGoogle didn\u2019t just have access to Plaintiffs\u2019 music; it operated the infrastructure through which much of that music reached the world.\u201d<\/p>\n While the earlier lawsuits against startups Suno and Udio targeted their music generators, the latest case against Google highlighted Google\u2019s \u201cstructural leverage\u201d as it doesn\u2019t just operate a music generator. It also owns YouTube<\/b> and runs Content<\/b> ID, YouTube\u2019s rights-management system that artists and labels rely on to police unauthorized use of their work.<\/p>\n The artists\u2019 lawyers wrote: \u201cGoogle didn\u2019t just have access to Plaintiffs\u2019 music; it operated the infrastructure through which much of that music reached the world. Google owns YouTube, one of the most important platforms for music discovery, distribution, and monetization, and it runs Content ID, YouTube\u2019s rights management system.\u201d<\/p>\n \u201cGoogle had every opportunity to develop this product legally\u2026 It has the technical infrastructure, financial resources, and industry connections to clear rights before training.\u201d<\/p>\n \u201cIts database contains over 50 million <\/b>reference files: full-length audio recordings provided by rights holders, including metadata identifying the artist, the rights owner, and the territories of ownership.\u201d<\/p>\n The plaintiffs argue that gave Google a structural leverage over artists, labels, and publishers. \u201cGoogle had every opportunity to develop this product legally\u2026 It has the technical infrastructure, financial resources, and industry connections to clear rights before training.\u201d<\/p>\n Google launched Lyria 3 on February 18 through its Gemini app, making the model available to more than 750 million<\/b> monthly active users<\/a>. Lyria 3 generates up to 30-second audio clips complete with vocals and lyrics. Users can also upload photos or videos and have Gemini compose an accompanying soundtrack.<\/p>\n In a blog post<\/a> announcing the launch, Jo\u00ebl<\/b> Yawili<\/b>, Senior Product Manager at Gemini App, and Myriam<\/b> Hamed<\/b> Torres<\/b>, Senior Product Manager at Google DeepMind, said: \u201cThe goal of these tracks isn\u2019t to create a musical masterpiece, but rather to give you a fun, unique way to express yourself.\u201d<\/p>\n Google has not specified how Lyria 3 was trained, but said in the same blog post that it has sought to \u201cdevelop this technology responsibly in collaboration with the music community\u201d and has \u201cbeen very mindful of copyright and partner agreements\u201d in training the model.<\/p>\n MBW understands that to mean the training for Lyria 3 uses music that YouTube and its parent company, Google, \u201chave the right to use\u201d under their \u201cterms of service, partner agreements, and applicable law\u201d.<\/p>\n The plaintiffs disputed that framing, pointing to Google\u2019s own published research<\/a> in 2022, in which Google researchers described how they gathered \u201cabout 50 million<\/b> internet music videos, extracting 30-second audio clips from each, and keeping roughly 44 million <\/b>clips after filtering, representing nearly 370,000 hours <\/b>of recorded music.\u201d<\/p>\n The indie artists also cited a 2023 paper<\/a> describing a separate training on 5 million<\/b> clips<\/b> totaling 280,000<\/b> hours<\/b> of audio.<\/p>\n \u201cBoth papers were published in peer-reviewed venues under the names of Google researchers. Neither mentions a license. Neither discusses a consent mechanism. Neither identifies a single rights holder whose permission was sought,\u201d according to the lawsuit.<\/p>\n The plaintiffs claimed that Google \u201cknew these practices created legal risk\u201d before it released any product to the public. In 2023, Google finished MusicLM<\/b>, but initially withheld the model from public release as the company acknowledged concerns about cultural appropriation and misappropriation of creative content, the court filing said.<\/p>\n As MBW previously reported, MusicLM became available for the public in May 2023. The tool worked by typing in a prompt like \u201csoulful jazz for a dinner party\u201d. The MusicLM model then created two versions of the requested song for the person inputting the prompt. Users can then vote on which one they prefer, which Google at the time said will \u201chelp improve the AI model\u201d.<\/p>\n The lawsuit also noted that Udio<\/strong> \u2014 which the plaintiffs (except for Kogon) previously sued for copyright infringement \u2014 was formed by four researchers at Google DeepMind: David Ding, Charlie Nash, Conor Durkan, <\/b>and Yaroslav Ganin.<\/b><\/p>\n \u201cWithin months, all four left Google DeepMind, formed<\/a> a new company, and publicly launched Udio, another AI music-generation service trained on copyrighted recordings without authorization,\u201d the lawsuit said.<\/p>\n In addition to their lawsuit against Udio, the AI startup also faced a lawsuit<\/a> from the RIAA<\/a><\/b> on behalf of the major labels. After Udio settled with Universal<\/strong> Music<\/strong> Group<\/strong><\/a> and Warner Music Group<\/strong><\/a>, Google launched<\/a> Lyria 3, the plaintiffs said.<\/p>\n