Distributor Symphonic has signed a partnership with start-up ArtyShield, putting a suite of anti-AI tools inside SymphonicMS, the dashboard its clients already use to manage releases. Artists can open the tools with a single sign-on and start free, with paid tiers for more coverage. The deal is aimed at independent artists and labels who distribute through Symphonic and want protection built into the release process instead of bolted on later.
The main product, MusicShield, treats a recording before it goes out. It alters features that machines pick up but listeners cannot hear, so that a model trained on the file learns little and produces degraded output. It grew out of HarmonyCloak, a 2024 research project led by ArtyShield founder Jian Liu at the University of Tennessee, and works on the same principle as Glaze and Nightshade for images. The research was presented at the IEEE Symposium on Security and Privacy, and the pitch is prevention rather than enforcement after the fact. MBW connects the launch to reporting that leaked Suno code listed YouTube Music, Deezer and Genius among its scraped sources.
The package also includes VeriTune, which estimates whether a track is AI-generated, VoiceShield against voice cloning, and VeriVoice for spotting synthetic speech. A certification service meant to prove a work is human-made is on the way, along with catalogue-wide infringement detection that can catch sped-up, pitch-shifted or AI-remixed copies and soundalike vocals.
Both chief executives framed the deal around consent: protection should come before a song is scraped, not after. Symphonic has made similar moves already, working with Warner-owned Sureel on tracking and opt-outs and with SourceAudio on a licensed training-data marketplace.
Sources · 1