The Atlantic's investigative team uncovered four massive datasets of music being used to train artificial intelligence models and created searchable databases for public access. Two of the datasets contain twelve million and nine million tracks respectively, with two additional datasets also holding millions of songs. The searchable databases allow musicians and the public to investigate whether specific songs were included in the training data, offering rare transparency into the information sources behind major AI systems.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton