“Neither Suno nor Udio have made their training datasets public. And both firms are vague about the sources of their training data — though that’s par for the course in the AI industry. (OpenAI, for example, has
dodged questions about whether YouTube videos were used to train its Sora video model.)”
—SC