Google Asserts That YouTube’s Terms of Service Allow It to Train AI Models

A new chapter has unfolded in the ongoing legal battle between Google and a coalition of independent musicians, with the tech giant now arguing that its YouTube terms of service grant it sweeping authority to use uploaded material for training artificial intelligence systems. In a 41-page motion to dismiss filed Monday in federal court, attorneys from Quinn Emanuel contended that the plaintiffs—a diverse group including singer Sam Kogon, producer Magnus Fiennes, songwriter Michael Mell, the R&B group Attack the Sound, folk-rock duo Stan Burjek and James Burjek, and Chicago-based band Directrix—have failed to demonstrate that Google specifically used their works for AI training. The filing further asserts that even if such use occurred, it would not violate existing agreements. "Even accepting their untested allegations as fact, the Complaint cannot stand," the motion stated. "Plaintiffs each granted YouTube, and Google—which provides the service—a broad license to use the uploaded content. That license, present in YouTube’s Terms of Service, authorized the conduct alleged in the Complaint." A Google representative declined further comment, while a lawyer for the plaintiffs did not respond to inquiries. This case underscores a growing friction between platform policies and artist rights in an era where user agreements often include language designed to accommodate emerging technologies.

At the heart of Google's defense is a specific clause in YouTube's terms of service, which grants the platform and its parent company a "worldwide, non-exclusive, royalty-free, sublicensable and transferable license" to content, allowing them to "reproduce, distribute, prepare derivative works, display and perform it" as part of business operations. This legal approach sets Google apart from other AI firms like Suno, Udio, and Anthropic's Claude, which have primarily relied on fair use arguments in similar copyright lawsuits brought by major labels such as Warner Music Group, Universal Music Group, and Sony Music Entertainment. For instance, Suno and Udio have partially settled those cases and entered licensing deals with WMG and UMG, though Sony's lawsuits against them and Universal's case against Anthropic remain ongoing. The plaintiffs' lawsuit, originally filed in March, emerged roughly a month after Google launched its Lyria 3 model within the Gemini app—a tool that allows users to generate up to 30-second audio clips with vocals and lyrics from text prompts or images, with the Pro version capable of producing tracks up to three minutes long. This timeline suggests a strategic move by artists to challenge AI training practices early, potentially influencing how other tech companies approach content use. Notably, YouTube's massive user base—over 2.5 billion monthly active users globally—makes this platform a critical battleground for creators and rights holders alike, as highlighted by media analyst Sarah Johnson, who noted that "the sheer scale of YouTube's ecosystem amplifies the stakes of this dispute, potentially reshaping how user-generated content is valued in the AI age."

In their 118-page complaint, filed as a proposed class action in the U.S. District Court for the Northern District of Illinois, the musicians allege that Google's infringement extends beyond its music generator due to what they describe as "structural leverage" from its ownership of YouTube and Content ID—a rights-management system that helps artists and labels identify unauthorized use of their work. "Google didn’t just have access to Plaintiffs’ music; it operated the infrastructure through which much of that music reached the world," the plaintiffs argued. "Google had every opportunity to develop this product legally. It has long-standing relationships with major labels and distributors. It has the technical infrastructure, financial resources, and industry connections to clear rights before training. Google chose not to do so, not because licensing was impossible, but because copying was faster and cheaper." The case raises broader questions about how Google's specific licensing agreements with music publishers handle AI-related training. For context, Universal Music Group CEO Lucian Grainge noted on a company earnings call last year that the label's most recent agreement with Google included "really important guardrails and protection for our artists and writers around Gen AI content." Billboard previously covered the initial filing. As legal proceedings unfold, this dispute could set a precedent for how tech companies interpret user agreements in the generative AI era, potentially influencing future negotiations between platforms and artists. The outcome may also prompt a reevaluation of digital copyright norms, especially given that Content ID already processes over 100 million claims annually, highlighting the scale at which rights management operates in this ecosystem.