Musicians claim Google trained its AI on their music. Google’s response? You already gave us permission.
Google is asking a court to throw out a copyright lawsuit filed by a group of independent musicians over its AI music model, Lyria.
The company’s argument is a simple one: if artists uploaded their music to YouTube, they already granted Google the rights it needed to use that music for AI training.
According to a new court filing, Google says the musicians involved in the case agreed to YouTube’s terms of service, which grant YouTube and its affiliated companies a broad licence to use uploaded content in a variety of ways.
Google argues that licence covers the conduct at the centre of the lawsuit – namely, using music uploaded to YouTube to help train Lyria, the company’s AI-powered music generation model.
The musicians behind the lawsuit claim their work was used without permission to develop technology that could eventually compete with the very artists whose music helped create it.
Google disagrees.
In its filing, the company points to a section of YouTube’s terms that grants a “worldwide, non-exclusive, royalty-free, sub-licensable and transferable licence” to use uploaded content, including the right to reproduce it and create derivative works.
Google’s lawyers also argue that YouTube’s parent company, Alphabet, and its various divisions fall under the definition of affiliated companies covered by those terms.
If that interpretation holds up in court, it could have major implications for independent musicians who upload content directly to YouTube.
Rather than relying on the increasingly common argument that AI training qualifies as fair use under US copyright law, Google is effectively saying it already secured permission through YouTube’s existing user agreement.
The case also highlights a growing frustration among artists over transparency in AI development.
Google’s filing notes that the lawsuit is based on the assumption that the musicians’ specific works were used to train Lyria.
The company argues there is no direct evidence proving that claim.
Critics, however, point out that artists have little way of knowing whether their music was included in AI training datasets because companies developing generative AI tools rarely disclose exactly what material was used.
That lack of transparency has become a central issue across the creative industries as musicians, authors and visual artists increasingly challenge how AI companies gather and use training data.
The dispute is particularly awkward for YouTube, which has spent years positioning itself as a creator-friendly platform while simultaneously expanding its investment in AI technology.
Google has previously said it wants to work alongside the music industry to develop AI responsibly.
For many artists watching this case unfold, the company’s latest legal argument appears to send a very different message.
The lawsuit is still in its early stages, and it remains to be seen whether the court agrees with Google’s interpretation of YouTube’s terms of service.
Either way, the case could become one of the most significant tests yet of how much control artists really have over the content they upload online.
