This guy developed an amazing way to push back against scraping original music to train AI models. Basically he adds noise to the track that isn't heard by the human ear, but really messes around with the models. The music being generated turns to sh$t! Lol
He also has a "mute attack" which he demonstrated by calling out to Amazon Alexa multiple times, but the device was unable to respond. Outside of music something like this would be amazing for stopping big tech eavesdropping on our conversations.
Well worth a watch!