Full Breakdown
Suno Hack Reveals Massive Music-Scraping Operation Behind AI Generator
7/17/2026, 2:01:12 AM
Core Event
In November 2025 a supply-chain attack—identified by the hacker as “ellie.191” and using the Shai-Hulud worm—compromised Suno’s internal systems. The breach gave the attacker access to source-code from 2023-2024, detailed scraping scripts, and a customer database containing email addresses, phone numbers and Stripe payment details. Suno described the incident as “limited” and said the exposed code was “outdated source code that is no longer in use at Suno.”
Background & Context
Suno, one of the largest AI-music platforms, has been sued by the Recording Industry Association of America (RIAA) and a coalition of major labels (Universal Music Group, Sony Music Entertainment, Capitol Records, Atlantic Records, Warner Music) for allegedly training its models on copyrighted recordings without permission. In a 2024 filing Suno admitted its training data “includes essentially all music files of reasonable quality that are accessible on the open Internet,” and it has argued that such use is protected by the fair-use doctrine. Warner Music settled its lawsuit in November 2025 and entered a licensing partnership with Suno; other cases remain pending.
Data & Statistics
Leaked files enumerate the scale of Suno’s data collection:
- YouTube Music – 2,013,545 clips; 113,879 hours of “youtube_music” plus 152,162 hours of “ytm_tagged.”
- Genius – 17,615 hours of “genius_hq.”
- Deezer – 12,287 hours.
- Pond5 – 62,117 hours (stock library owned by Shutterstock).
- International Music Score Library Project (IMSLP) – 19,514 hours.
- Freesound – 410 hours; Jamendo – 3,726 hours; MuseScore lyrics – 103 hours.
- Podcasts – 420,000 titles identified via PodcastIndex, with code aiming to download roughly 1 million hours of audio.
The code also shows Suno used Bright Data proxies to “stream-rip” YouTube tracks and specifically searched for a-capella versions to isolate vocals.
Official Statements & Responses
Suno’s spokesperson told media that the breach was quickly contained, that “no sensitive personal information was compromised,” and that Suno “does not have access to customers’ full credit card numbers in Stripe.” The company added that, because the exposed data involved only “outdated source code,” it deemed individual breach notifications “not warranted under applicable privacy laws.”
Regarding its training practices, Suno reiterated that its models are built on “publicly available music files and related metadata accessible on third-party websites on the open Internet” and emphasized a design philosophy of “Original Creation, By Design,” which excludes artist names from training metadata and incorporates detection filters to block prompts that replicate existing works.
Criticism & Opposition
The RIAA’s amended complaint alleges Suno “unlawfully ‘stream ripping’” YouTube recordings and circumvented the platform’s anti-scraping encryption, seeking statutory damages of up to $150,000 per work and $2,500 per act of circumvention.
Artists For An Ethical and Sustainable Internet launched the “Say No To Suno” campaign, describing the platform as a “brazen ‘smash and grab’” operation that profits from artists’ work without compensation.
Deezer issued a statement that “training data for generative AI should always be attained with clear approval and fair compensation,” and said it is “assessing the situation and considering available options.”
Conflicting Reports & Gaps
Several customers contacted by 404 Media confirmed they never received breach notifications, contradicting Suno’s claim that notification was unnecessary. While Suno asserts the exposed code is obsolete, the leaked files date from 2023-2024 and contain active scraping instructions, leaving the extent of current data-collection practices unclear.
Verbatim Quotes
- “As we have stated in public filings and disclosures, Suno’s AI models have been trained on publicly available music files and related metadata accessible on third-party websites on the open Internet.” — Suno spokesperson
- “Suno obtained those copies in the first instance by unlawfully ‘stream ripping’ them from the popular streaming platform YouTube, and circumventing the technological measures designed specifically to prevent such unauthorized copying.” — RIAA (lawsuit)
- “Our goal has always been to help people create original new music, not replicate someone else’s. That’s why we build our models around what we call ‘Original Creation, By Design.’ For example, we intentionally do not use artist names as a category of training metadata because we want our models to help people create brand new songs, not music that replicates other artists’ existing work,” — Suno spokesperson
- “I like to hack anything and everything.” — ellie.191, the hacker who supplied the data
- “Based on the limited nature of the customer information believed to be involved, we determined that individual notifications were not warranted under applicable privacy laws.” — Suno spokesperson
What’s Next
The RIAA-led lawsuit remains before a federal judge, who has yet to rule on the statutory-damage claim that now covers over 61,000 identified works. Suno continues to operate while defending its fair-use argument and refining its detection safeguards.
