Jamendo Sues Nvidia for $20M Over AI Music Training Data

Nvidia is facing its eighth copyright lawsuit related to AI training data. Luxembourg-based music platform Jamendo has filed a federal suit, alleging Nvidia used its audio dataset without authorization.

The complaint, filed in the US District Court for the Northern District of California, seeks a minimum of $20.3 million in damages. Statutory damages could reach up to $150,000 per infringed work.

At issue is the MTG-Jamendo dataset, a collection of approximately 55,000 audio tracks developed for non-commercial research. Jamendo alleges Nvidia used these tracks to train its commercial AI models, Fugatto and Audio Flamingo, violating the dataset's Creative Commons license.

This lawsuit highlights the legal risks surrounding AI training data. Unlike text-based disputes, audio datasets often have clearer licensing terms, making fair use arguments more difficult. The cumulative financial exposure from multiple lawsuits is becoming significant for Nvidia.