After Anna's Archive published a massive scrape containing 86 million songs and metadata from Spotify, the streaming giant responded by disabling the nefarious accounts responsible. A spokesperson for Spotify told Recorded Future News that it "has identified and disabled the nefarious user accounts that engaged in unlawful scraping." "We've implemented new safeguards for these types of anti-copyright attacks and are actively monitoring for suspicious behavior," the spokesperson said. "Since day one, we have stood with the artist community against piracy, and we are actively working with our industry partners to protect creators and defend their rights." The Record reports: The spokesperson added that Anna's Archive did not contact them before publishing the files. They also said it did not consider the incident a "hack" of Spotify. The people behind the leaked database systematically violated Spotify's terms by stream-ripping some of the music from the platform over a period of months, a spokesperson said. They did this through user accounts set up by a third party and not by accessing Spotify's business systems, they added. Anna's Archive published a blog post about the cache this weekend, writing that while it typically focuses its efforts on text, its mission to preserve humanity's knowledge and culture "doesn't distinguish among media types." "Sometimes an opportunity comes along outside of text. This is such a case. A while ago, we discovered a way to scrape Spotify at scale. We saw a role for us here to build a music archive primarily aimed at preservation," they said. "This Spotify scrape is our humble attempt to start such a 'preservation archive' for music. Of course Spotify doesn't have all the music in the world, but it's a great start."

Read more of this story at Slashdot.

A group of activists has scraped Spotify's entire library, accessing 256 million rows of track metadata and 86 million audio files totaling roughly 300TB of data. The metadata has been released via Anna's Archive, a search engine for "shadow libraries" that previously focused on books. Spotify described the activists as "anti-copyright extremists who've previously pirated content from YouTube and other platforms" and confirmed it is actively investigating the incident. The activists claim this represents "the world's first 'preservation archive' for music which is fully open" and covers "around 99.6% of listens." They appear to have used Spotify's public web API to scrape the metadata and circumvented DRM to access audio files. Spotify insists that this is not a security breach affecting user data. Though the more pressing concern for the music industry may be AI training rather than pirate streaming services -- similar YouTube datasets have reportedly been used by unlicensed generative AI music services.

Read more of this story at Slashdot.

An EDM song by the British group Haven ran into trouble in October after it shared clips of upcoming song "I Run" on TikTok. The song "was an overnight viral sensation online," writes Digital Music News — racking up millions of plays "even before it hit streaming services." (Although the Washington Post notes that "Record labels and TikTok users began questioning whether 'I Run' used an AI deepfake, modeled off British R&B singer Jorja Smith, for the vocals.") Digital Music News picks up the story: The artist says he used his own voice to record the vocals, and then ran it through layers of processing and filtering to turn it into the female-sounding voice heard in the track. However, that filtering also included the use of the controversial genAI platform Suno — and that's what complicates things... [The article says later that Suno "is currently in the middle of a blockbuster lawsuit with the Big Three major labels over allegations of widespread copyright infringement of sound recordings used during the AI model training process."] Meanwhile, the song was rapidly amassing listenership. It soared to #11 on the U.S. Spotify chart and #25 on Spotify globally. Videos using the song continued going viral on TikTok and Instagram, including one in which rapper Offset had apparently played the song during a Boiler Room set, which later turned out to be falsified. And then, as quickly as it appeared, "I Run" was taken down from streaming services, including Spotify and Apple Music. That was due, in part, to numerous takedown notices from The Orchard, the label to which Jorja Smith is signed, as well as the RIAA and IFPI. The takedown notices alleged various issues with the track, including the "misrepresentation" of another artist, as well as copyright infringement. As a result, the song has also been withheld from the Billboard charts, including the Hot 100, on which it had been predicted to debut this week before the controversy. Billboard points out that it "reserves the right to withhold or remove titles from appearing on the charts that are known to be involved in active legal disputes related to copyright infringement that may extend to the deletion of such content on digital service providers." The song itself has now been re-released with an all-human vocal track. But going forward will the music industry ever work with AI platforms? The Washington Post reports: "I Run" has taken off as record labels remain unsure of the extent to which they should welcome generative AI programs such as Suno or Udio into the industry. After the two AI music companies began growing in popularity, the three major labels — Sony Music, Warner Music Group and Universal Music Group — filed lawsuits against Suno and Udio, claiming that the AI companies have used the labels' sound recordings to train their model. Since then, UMG and Warnerhave reached agreementsto work with Udio, ending their litigation... It comes shortly after all three major labels licensed their catalogue to Klay, a music streaming start-up that allows users to adjust songs using artificial intelligence. Major licensing organizations such as ASCAP and BMI shared that they would register songs that were partially AI-generated — but not fully generated ones. Haven appears to present an uncomfortable edge case. While some AI-generated songs that sound broadly like other artists have been allowed to remain on streaming platforms, the voice in "I Run" appears to have been deemed too duplicative for comfort.

Read more of this story at Slashdot.