Hacked source code reveals how Suno scraped decades worth of music and podcasts from the internet to train its AI tool.

The code also suggested that Suno was using proxies to scrape songs from YouTube through a company called Bright Data, which sells scraping tools, infrastructure, and data services. Additional code shows that with the help of an online tool called PodcastIndex, Suno identified 420,000 different podcasts that had at least five, 30-minute episodes and sought to download roughly 1 million hours of podcasts

The hacker, ellie.191, told 404 Media they breached the company by hacking an individual employee using the Shai-Hulud worm, a supply chain attack that allowed hackers to harvest GitHub and cloud service credentials. They said they also accessed Suno’s customer list, which included customers’ emails and/or phone numbers and Stripe payment details, depending on what they used to login. The hacker provided a sample of some of the customers, some of whom confirmed to 404 Media they had used their phone number to sign up for Suno and said they were never notified of a breach.

Last month, The Atlantic reported on several music databases that are widely used in AI training, consisting of millions of tracks: “Three of the datasets I found are distributed as a list of links to songs on YouTube or Spotify. AI developers download the actual audio using tools that automate the job, some of which allow developers to bypass logins, advertisements, and mechanisms that might earn money or subscribers for creators. Such tools violate the terms of service of these platforms.

Archive link: https://archive.ph/xX3XW

  • Grimy@lemmy.worldEnglish
    9·
    9 hours ago

    We must band together and protect the record companies! All music generation must go through Universal and Warner’s proprietary app. Death to training without paying the stockholders, death to open-source.

    All hail censorship, all hail Google.

    • Telorand@reddthat.comEnglish
      5·
      5 hours ago

      I personally know indie musicians who are popular enough to have had their music scraped and used as training data, in part because they are forced to use platforms like Spotify, YouTube, etc. to eke out a living. And then people have had the audacity to present AI slop versions of music using their vocals to them, expecting them to be happy about it.

      Were they paid for their decades of work? No. Were they given the opportunity to opt out? No. This isn’t The Man trying to capitalize upon an open and equitable system. This is The Man stealing from the poors and each other, and pretending they didn’t (or worse, acting like their victimhood is somehow equal to everyone else’s).

      And to add even further insult to injury, these companies have tricked many of those same poors that these plagiarism engines are tools to “level the playing field” against people who have spent years honing their craft. Why, the very nerve of those people using their time to become experts! /s

      Death to training without paying the stockholders, death to open-source…

      Even better: how about death to training until you pay the original artists fairly and equitably, or allow them to opt out? AI should not be treated as some inevitable thief that we must all work around.

      • Grimy@lemmy.worldEnglish
        3·
        5 hours ago

        Fair points but the current lawsuits arent there to solve any of them. The lawsuit against Udio essentially gave ownership of the app to UMG and Warner. No artists were payed. The app still exists, except you have to pay to download the song and they put extra copyrights on every song generated now.

        It’s insanely transparent imo, they want a monopoly on generated music as well as the regular stuff. I wish artists had a chance.

        Even worst is that UMG will probably start pushing AI music once the dust settles because there’s more profit in it, especially if they own it the moment it’s made and can dictate the terms with a simple ToS.

      • Grimy@lemmy.worldEnglish
        1·
        4 hours ago

        All the current lawsuits kind of suck because its all copyright juggernaut vs big AI companies. It feels mostly lose lose for us but only one of the two outcomes kneecaps open source and community driven options.

        Another frustrating bit is that it’s win win for big AI. Even if they lose, it just means a cheap fine while getting a massive moat because of the new cost of entry.