Adobe adds music generation to Firefly and Premiere
Firefly can now turn a prompt or video into a track, while Premiere's beta places the same workflow directly inside an edit.
What’s happening in AI music
Important releases, creative work, rights decisions and industry shifts — explained clearly and linked to the original sources.
Latest stories
One detector classified 38.5% of July releases as fully or partly AI-made
SubmitHub says its SH Labs system scanned more than one million releases. The result is a platform estimate, not an audited measure of all music released worldwide.
Apple Music is preparing visible labels for AI-generated content
Apple's delivery specification already has structured AI metadata. Providers have now been told that public-facing labels are coming later in 2026.
An AI album appeared under composer Kumi Tanioka’s name
The music was not hers. The case exposes a distribution and metadata failure that makes synthetic releases look official.
Udio and GRAI sign the first licences for UMG's AI-music patent portfolio
Music IP Holdings is offering a single licence across more than 24 issued or allowed patents, with another 50-plus applications pending.
More to know
Spotify will label AI-generated artist identities and keep them out of recommendations by default
AI Persona badges start appearing in September. Spotify says it will review profiles as well as accept self-disclosure, and will not recommend those artists unless listeners follow them.
Tamber 2.0 uses AI to help musicians control the tools they already have
The updated suite adds Windows support, gesture-controlled instruments and a conversational layer for Ableton without generating finished songs.
AI music generators do not need to replace the DAW
The valuable frontier is not a browser copy of Ableton or Logic. It is better control over what enters a model, what can be regenerated, and what leaves as editable material.
ElevenLabs gives Music v2 reference audio and section-level control
The current ElevenMusic model can build and revise a song in sections, use reference audio and repair selected passages instead of regenerating the whole track.
HappyShrimp targets full-song coherence with hierarchical planning and flow matching
The new song product arrives beside an Alibaba Token Foundry technical report built around global planning, local token generation and full-song audio rendering. The public paper does not itself name HappyShrimp.
Pika launches a full-song music model for its API
Pika Music combines prompts, lyrics and optional voice or music references, produces tracks up to six minutes, and is priced by successful output minute.
Suno Studio 2.0 adds MIDI, programmable effects and a conversational production layer
Suno's browser studio can now record and edit MIDI, turn clips into generation prompts, automate effects and build account-level tools through chat. Access remains Premier-only.
Treblo makes Melodia v3 free and unlimited — and releases a detector for its own songs
The former Sonauto is offering its latest consumer model without generation credits or daily caps. The API remains metered, and its open detector is deliberately Treblo-specific.