SZA Torches Suno Over 238 Scraped Tracks

<p>SZA dropped a flamethrower on generative AI music platforms June 22 after discovering 238 of her songs, including unreleased material, had been scraped for training without consent or payment.</p> <h2>🎤 The IG Manifesto</h2> <p>In a series of raw Instagram posts, the Grammy-winning artist didn’t mince words. “If you’re a musician and you support this degenerate shit? You’re disgusting,” she wrote, directly naming Suno and calling out Diplo for his investment in a related AI startup. She framed the issue as cultural theft layered with environmental racism, noting AI data centers disproportionately harm Black and Brown communities while profiting off their creativity.</p> <p>SZA highlighted the overrepresentation problem: Black artists make up just 13% of the population but dominate training data for music models. Her directive was clear—"DO NOT GIVE AWAY YOUR VIBRANIUM !!! DO NOT TRAIN AI W YOUR GENIUS." The posts quickly went viral, amplifying existing artist fury and putting Suno back in the crosshairs.</p> <h2>⚖️ Sony's Active Lawsuits</h2> <p>This isn’t abstract. Sony Music, which distributes SZA via RCA, has been suing Suno and competitor Udio since 2024 for copyright infringement. Suno has admitted in court filings to training on “tens of millions of recordings” and continues to argue fair use. The Atlantic’s recent AI Watchdog investigation revealed the staggering scale—datasets with 12 million, 9 million, and hundreds of thousands of tracks circulating in AI dev circles.</p> <p>Nicole Atkins separately discovered 41 of her copyrighted songs in Suno’s training pool via the same reporting. No permission. No royalties. The pattern is clear: platforms launched with polished marketing about “democratizing music” while building on uncompensated work from professionals.</p> <h2>📈 What It Means for AI Creators</h2> <p>For producers and artists using Suno, Udio, and similar tools professionally, the backlash is shifting the landscape. Distributors are tightening AI-content rules. Playlists are rejecting fully machine-generated tracks. Labels are watching every viral AI hit as potential litigation fodder.</p> <p>Meanwhile, SZA’s call for Black musicians to withhold their styles could fragment the training data pools that make these platforms sound good in the first place. Expect more public artist audits, more lawsuits, and pressure on platforms to implement opt-out mechanisms that actually work.</p> <div class="takeaway"><p><strong>Bottom line:</strong> SZA just made ignoring consent in AI music training culturally radioactive—platforms and users who don’t adapt will face escalating boycotts and legal heat.</p></div>