Suno snatched millions of songs from YouTube, Genius, and Deezer

Suno information obtained successful a hacking incident has exposed that the AI euphony generator was trained by scraping millions of songs and lyrics from online audio platforms, including YouTube Music, Deezer, and Genius, 404 Media reports. Given that Suno has avoided revealing what’s successful its training datasets and really they were acquired, this a uncommon glimpse into what Suno has actually been taking from online platforms.

That’s applicable because Suno has been the taxable of respective lawsuits that allege it utilized copyrighted materials to train its AI models. In a notable case revenge by the Recording Industry Association of America (RIAA), Suno openly admitted that it does so, arguing that training connected copyrighted materials and publically disposable euphony files from the unfastened net is legally permitted nether adjacent usage doctrine. Whether aliases not a tribunal agrees pinch that, an amendment revenge by the RIAA past twelvemonth besides alleges that Suno unlawfully circumvented YouTube’s copyright protections by intentionally “stream ripping” tracks from the platform.

Materials shared pinch 404 Media by the hacker, referred to arsenic “ellie.191,” reportedly backmost up those allegations. The information includes Suno root codification from 2023 and 2024, alongside scraping instructions to propulsion audio files from YouTube Music, Deezer, Genius, Pond5, Jamendo, Freesound, and the International Music Score Library Project (IMSLP). Other leaked codification reportedly suggests that Suno utilized a third-party institution called Bright Data to scrape euphony from YouTube, and seemingly searched for a cappella versions of songs connected the level to root vocal-only audio.

A record for YouTube Music notes that Suno had consumed 2,013,545 YouTube Music clips astatine the constituent it was past updated. According to different file, datasets compiled by Suno included hundreds of thousands of hours of YouTube Music, thousands of hours of Deezer, Genius, IMSLP, Jamendo, and Pond5, and hundreds of hours of Freesound and MuseScore lyrics. Suno besides sought to download astir 1 cardinal hours of podcasts via an online instrumentality called PodcastIndex, according to further code.

“As we person stated successful nationalist filings and disclosures, Suno’s AI models person been trained connected publically disposable euphony files and related metadata accessible connected third-party websites connected the unfastened Internet,” an unnamed Suno spokesperson said successful a connection to 404 Media.

Suno customer accusation was besides accessed by the hacker, which included email addresses, telephone numbers, and Stripe costs details. Some of the customers contacted by 404 Media confirmed that they had signed up for the service, and said that Suno ne'er notified them astir a information breach.

In a connection to 404 Media, a Suno spokesperson said the institution became alert of a information incident successful November 2025, and that the business was quickly contained.

“At the time, we instantly conducted an investigation and verified that the incident chiefly progressive outdated root codification that is nary longer successful usage astatine Suno and that nary delicate individual accusation was compromised. Importantly, Suno does not person entree to customers’ afloat in installments paper numbers successful Stripe,” the Suno spokesperson said. “Based connected the constricted quality of the customer accusation believed to beryllium involved, we wished that individual notifications were not warranted nether applicable privateness laws.”