Overview
- On June 10, Google filed a 41-page motion to dismiss arguing that YouTube’s terms of service give the company a broad, royalty-free license to reproduce and “prepare derivative works” from uploads and that the plaintiffs cannot show Google used their specific files.
- The proposed class-action complaint, filed in March by independent musicians including Sam Kogon, Magnus Fiennes and the bands Directrix and Attack the Sound, says Google trained its Lyria 3 music model on creators’ YouTube uploads without consent or payment and exploited its control of YouTube and Content ID.
- Google has not publicly confirmed that specific uploads were used to train Lyria 3, but past YouTube statements acknowledge using uploaded content to improve and train other Google models and Google told the court it trained Lyria on material it had a right to use.
- Plaintiffs allege large-scale scraping—reported as roughly 44 million clips in their papers—and a court denial of Google’s motion would likely open discovery so artists can seek evidence tying particular uploads to the model.
- A ruling for Google would strengthen the use of platform contracts as a defense and could let AI companies rely on user terms instead of licensing, while a ruling for artists could force broad licensing deals or new data-sourcing strategies and affect how creators are paid.