AI-Generated Music Tools: Creating Cloned Artist Voices and Turning Humming into Melodies on YouTube

YouTube unveiled new experimental AI services on Thursday, including Dream Track in YouTube Shorts, generating up to 30-second soundtracks using AI-generated artists’ voices. Musicians pushed back on AI using their voices without permission, but YouTube got nine big names to participate, including John Legend, Troye Sivan, CharliXCX, and T-Pain. However, the feature has been tied up in negotiations over rights and payments with recording companies.
Users can access Dream Track by typing an idea and choosing from participating artists, using Google DeepMind’s Lyria, designed for creating high-quality vocals and instrumentals, with all content marked with a SynthID watermark.
Some artists are apprehensive about AI but see collaborative opportunities. Music AI Tools are also coming to YouTube, creating guitar riffs, and transforming tracks. YouTube recently announced new policies for labeling AI videos and letting public figures report deepfakes. Dream Track is available to a select group, with Music AI Incubator participants able to test the tools later this year.