Suno, the AI music-making platform being sued by the record majors over copyright infringement, has once again unveiled upgrades to its platform\u2019s capabilities.<\/p>\n
The upgrades, announced on Suno\u2019s blog<\/a> on Tuesday (June 3), come amid news that Suno, along with rival AI music platform Udio<\/strong> (also facing a lawsuit by the music majors) are in talks with Sony Music Entertainment<\/strong>, Universal Music Group<\/strong> and Warner Music Group<\/strong> to license the music giants\u2019 recordings.<\/p>\n The latest news on that front came from the Wall Street Journal<\/em>, which reported<\/a> on Tuesday that among the key asks from the recording companies is a Content ID-style fingerprinting technology that would make it possible to track when and how a particular song is used by the AI platforms.<\/p>\n Content ID is YouTube<\/strong>\u2019s process for identifying music that\u2019s been uploaded to the video platform, enabling YouTube to distribute revenue from ads served on a song to the proper rights holders, regardless of whether or not the uploader had permission to post the song.<\/p>\n Content ID is widely credited with ending years of acrimony between YouTube owner Google<\/strong> and music rights holders over unauthorized uploads.<\/p>\n WSJ<\/em> also reported that the music companies want to be \u201cactive participants\u201d in determining which products the AI music companies develop, and how they work.<\/p>\n That follows an earlier report<\/a> from Bloomberg<\/em>, which said the three majors are seeking licensing fees from Suno and Udio, along with a \u201csmall amount\u201d of equity in both companies.<\/p>\n In the meantime, the AI platforms continue to develop new products and capabilities.<\/p>\n On Tuesday, Suno unveiled an upgraded version of its Song Editor tool, which enables users to upload work-in-progress tracks that can then be rearranged or remixed within Suno.<\/p>\n The Song Editor can also alter lyrics, and the maximum length of an uploaded track has been increased to eight minutes.<\/p>\n Users can also start a new track by humming a melody or typing in a text prompt. The Song Editor now includes three new \u201ccreative sliders,\u201d which allow users to determine the \u201cweirdness\u201d level of the track, how structured it is, and how \u201creference-driven\u201d it is.<\/p>\n Users can complete the track within the Song Editor, or separate it into as many as 12 stems (vocals, drums, bass, etc.) and transfer the stems to the digital audio workstation (DAW) of their choice.<\/p>\n In a sign of how quickly AI music-making technology is evolving, the upgrades come roughly a month after Suno unveiled version 4.5<\/a> of its platform, which included improvements to the range and emotional depth of AI-generated vocals, allowing users to \u201ccreate everything from delicate, intimate performances to powerful deliveries with vibrato.\u201d<\/p>\n The new version also includes the capability to switch a song\u2019s genre and mash up genres, creating, for example, tracks that blend EDM with folk music, or emo with neosoul. The maximum generated song length was increased to eight minutes.<\/p>\n Meanwhile, the copyright infringement lawsuits against Suno and rival Udio continue.<\/p>\n Record companies owned by Sony, Universal and Warner filed the suits last year<\/a>, alleging \u201cmass infringement of copyrighted sound recordings,\u201d and offering evidence that, when prompted, the music generators will spit out music and lyrics that are very similar, if not identical, to existing songs.<\/p>\n In a response to the lawsuits last August, Suno and Udio all but admitted<\/a> to having trained their AI models on copyrighted music, but argued that this should be seen as a \u201cfair use\u201d exemption to copyright laws.<\/p>\n Suno is also facing<\/a> a copyright infringement suit brought by German collection society GEMA.<\/p>\n