Google and Apple debut AI music: creation vs. curation
Google’s Gemini now creates original AI music with Lyria 3, while Apple’s Apple Music leans on “Playlist Playground” to build smart playlists from existing tracks.
Summary
Google and Apple are rolling out new AI-powered music features, but with very different philosophies. Gemini with Lyria 3 generates original tracks, while Apple Music’s Playlist Playground focuses on curating playlists from existing songs.
In practice
On Google’s side, the Gemini app can now generate 30‑second tracks using the Lyria 3 model from text prompts about genre, mood or personal meaning, and even from photos and videos. The tool writes lyrics, lets users control style, vocals and tempo, and embeds SynthID watermarks in the audio while also checking whether uploaded clips were created with Google’s AI.
Google stresses that Lyria 3 is designed for original expression, not copying specific artists, even when a name is mentioned in the prompt. The feature is rolling out first on desktop to users 18 and older and is being folded into YouTube’s Dream Track tool for creating soundtracks for Shorts.
Apple, meanwhile, has added a new Playlist Playground feature in the first iOS 26.4 developer beta inside Apple Music. Users type prompts about moods, genres or themes — such as “morning coffee music” or “disco songs that defined the 1970s” — and get a roughly 25‑song playlist they can then tweak.
Those playlists can be refined with follow‑up prompts and edited manually, including adding or removing tracks, reordering songs, and changing cover art and descriptions. Apple includes a disclaimer that Playlist Playground may produce unexpected results and notes that it relies on Apple Intelligence models that download in the background before becoming available.
Context
Lyria 3 positions Gemini alongside standalone AI music tools by giving users quick, original tracks inside a general‑purpose assistant that already handles text, images and video. For casual users and creators, that turns music making into an everyday, low‑friction activity.
Apple’s Playlist Playground, by contrast, keeps the focus on music discovery and personalization rather than synthetic songs competing with its catalog. That approach is likely more comfortable for record labels and artists, who see recommendation and curation as a safer use of AI than full song generation.
More broadly, AI music is in a phase of rapid expansion as platforms like Suno and Udio grow and major labels move from lawsuits to licensing deals. With Microsoft, TikTok, Google and Apple all in the mix, AI is becoming a standard layer across both creation tools and mainstream music apps.
Why it matters
- Reveals two competing visions for AI in music: generating new songs versus smarter use of existing catalogs.
- Brings AI music features into everyday apps like Gemini, YouTube Shorts and Apple Music, not just niche tools.
- Reflects a shift by major music labels from pure litigation toward licensing around AI music.
- Raises fresh questions over copyright, artist compensation and how far AI should go in creative fields.