Suno AI’s music generator scraped millions of copyrighted songs and podcasts without permission. The company pulled content from YouTube Music, Deezer, Genius, and various stock music libraries.
The scale is staggering. Suno ingested over two million music clips from YouTube Music alone, totaling 113,879 hours. Other platforms added tens of thousands more hours, including 17,615 from Genius and 12,287 from Deezer. The tool also targeted 420,000 podcasts, aiming to download roughly one million hours.
Suno didn’t just grab random files. Its source code reveals precise scraping instructions targeting vocals by searching for acapella versions on YouTube. They used proxies through Bright Data to bypass restrictions and engaged in “stream ripping” — copying music while evading technological barriers.
The company admitted its training data covers “essentially all music files of reasonable quality accessible on the open internet.” That includes decades of popular sound recordings copied unlawfully to train its AI models to mimic human performances.
Besides copyright violations, Suno exposed user data. Hackers accessed information for hundreds of thousands of customers, including Stripe payment details. Suno’s spokesperson claims no full credit card numbers were compromised and blames an outdated source code for the security breach.
This isn’t an isolated incident. Major publishers and authors are suing Google over similar copyright violations. Google allegedly trained its Gemini AI on millions of copyrighted books and works from Google Books and Google Play without permission. The lawsuits accuse Google of removing or altering copyright metadata to hide its training sources.
One lawsuit filed in the Southern District of New York alleges Google knowingly reproduced copyrighted works without compensation. It claims Gemini encourages copycat creations without credit or payment to original authors.
The legal fallout from AI training on stolen content is growing. Anthropic, another AI company, paid a historic $1.5 billion settlement for pirating copyrighted works. Around half a million writers qualified for at least $3,000 each. Many authors declined the settlement to pursue further action.
The Suno case exposes the raw reality behind AI music tools promising unlimited creativity. As Mauvis Ledford, CEO of Sogni AI, put it, “Unlimited creative AI has been difficult to sustain because generation costs rise directly with usage.” That includes the legal and ethical costs of training AI on stolen art.
AI companies keep pushing boundaries while brushing aside copyright law. The Suno hack and lawsuits against Google show the industry faces an inevitable reckoning. The question is how many more artists and authors will pay the price before the tide turns.
Based on
- Hack Reveals Suno AI Music Generator Scraped YouTube, Deezer, and Genius — 404media.co
- Today’s Vergecast: The problem with Suno and AI music. | The Verge — theverge.com
- As AI Platforms Move Away From Unlimited, Sogni AI Launches a $20 Fair-Use Unlimited Plan on Community GPUs | Currency News | Financial and Business News | Markets Insider — markets.businessinsider.com
- Three Publishers Challenge Google Over AI Copyright Infringement — engadget.com
- Google faces another AI training lawsuit from major publishers | TechCrunch — techcrunch.com
