Google Says Your YouTube Uploads Are Already Licensed for AI Training

Google insists that the act of training a generative AI model falls squarely within this license—specifically under rights to reproduce content and create derivative works—and therefore requires no additional permission or payment to artists . The company has urged the court to dismiss the complaint 'with prejudice,' meaning the lawsuit could not be refiled .

Google is walking a careful legal tightrope, simultaneously refusing to confirm publicly whether it actually used specific YouTube creator content to train Lyria 3 while arguing that its terms make such use perfectly legal .

Inside the Musicians' Lawsuit

The proposed class action was filed on March 6, 2026, in the US District Court for the Northern District of Illinois by a group of independent musicians and songwriters including Nick Kogon, James Burjek, Berk Ergoz, Hamza Jilani, Maatkara Wilson, Arjun Singh, Magnus Fiennes, and Michael Mell . The law firm Loevy & Loevy represents the plaintiffs .

The 118-page complaint alleges that Google copied millions of copyrighted sound recordings, musical compositions, and lyrics from YouTube and across the internet to build its AI music generator, Lyria 3 . The lawsuit contends that Google, which owns YouTube, used its unique position to pivot 'from distributor to competitor' . Key claims include:

• Mass infringement: Google allegedly scraped at least 44 million clips and 280,000 hours of music to train its model without authorization .
• Direct competition: Google launched Lyria 3 on February 18, 2026, integrating it into the Gemini chatbot app with over 750 million monthly active users. A week later, it introduced ProducerAI, a standalone platform for creating, sharing, and monetizing AI-generated music . The plaintiffs describe Google's AI music operation as a 'vertically-integrated syndicate that illegally copies music, launders ownership, and unfairly competes' with human musicians .
• Novel biometric claims: In a legal first for AI music litigation, the suit includes a claim under Illinois' Biometric Information Privacy Act (BIPA), alleging that Lyria 3's voice cloning and vocal analysis features illegally collect and exploit artists' 'voiceprints' .

What Happens If Google's Argument Wins

If a court accepts Google's interpretation of the YouTube Terms of Service, the implications would reach far beyond this single lawsuit and could reshape the relationship between creators and the platforms that host their work.

No opt-out or payment for creators. The most immediate consequence is that any musician, podcaster, or video creator uploading to YouTube would be deemed to have automatically consented to their work being used to train AI that may eventually compete with them, with no mechanism to revoke permission or demand compensation .

A blueprint for other platforms. Other major platforms with similarly broad licensing language in their terms—such as Instagram and TikTok—would have a clear roadmap to defend their own AI training practices against copyright challenges. This could make it significantly harder for creators to challenge unauthorized AI training industry-wide .

Undercutting future licensing deals. A ruling in Google's favor could gut the leverage creators currently have in pushing for opt-in AI training deals and separate licensing fees from AI developers. If platform terms are ruled sufficient authorization, the emerging market for AI training licenses could collapse before it matures .

Covering all content, not just music. The logic of Google's argument is not limited to songs. The same TOS clause applies equally to videos, podcasts, tutorials, livestreams, and any other creative work uploaded to YouTube, potentially giving the company free rein to use all user-generated content for future AI models .

The case remains pending in Chicago federal court. Its outcome will likely influence parallel lawsuits against other AI music companies and help define the boundaries of copyright in the age of generative AI.