Friends, after the advent of AI, things have become very easy, the biggest proof of this is the music made with AI, the ongoing legal battle over which has taken a new and shocking turn. Earlier this year, Google introduced its powerful music-generation model, Lyria 3, in Gemini, which lets users create songs using only text prompts. Soon after its launch, there were allegations that Google trained AI using songs uploaded to YouTube without paying the original creators.
Google’s response to the YouTube-AI Music case has sparked a major debate about the rights of creators.
Shortly after its release, a lawsuit accused the company of training models on music uploaded to YouTube without paying the artists or getting their explicit approval.
Google’s shocking answer
According to documents submitted to the court, Google argued that whenever artists upload their music to YouTube, they grant the company a license that allows them to use that content for a variety of purposes, including AI training. The company said YouTube’s Terms of Service allow such use, meaning content uploaded to the platform can legally be used to develop and improve AI systems.
Did Google Admit to Using YouTube Songs for AI Training?
Not completely.
Google did not explicitly confirm that Lyria 3 was trained using songs uploaded to YouTube. Instead, his legal team argued that the allegations made in the lawsuit are false.
If the court accepts Google’s interpretation, it could determine that the company has the legal right to train its AI models using music uploaded to YouTube—without any additional permission or payment.
How Google’s approach is different from other AI companies
Companies like Suno, Udio, and Anthropic have defended themselves primarily by arguing that training AI on copyrighted material is “fair use” because it creates transformative technologies that benefit society at large.
Comments are closed.