Spotify has reportedly pulled “tens of thousands” of Boomy songs following a days-long distribution pause resulting from alleged fake streams – though the sizable collection of removed works represents just seven percent of the AI music generator’s on-platform releases.
These newest details about the Boomy-Spotify impasse (and AI’s rapidly evolving industry role) emerged in a Financial Times piece, after the former platform revealed over the weekend that it would resume the “curated delivery” of its tracks to Spotify.
But in making the announcement, Boomy – which says it enables users to create “original songs in seconds” even if they’ve “never made music before” – opted not to disclose how many of its works were on the Stockholm-based service or how many of these tracks had been taken down. For reference, the AI technology at hand has since Monday of last week been used to produce an average of over 23,000 songs per day, according to data displayed on Boomy’s website.
Meanwhile, the aforementioned FT piece has indicated that Spotify delisted some seven percent of Boomy’s uploaded music, or tens of thousands of works, after Universal Music Group purportedly identified and reported (to all leading platforms) suspicious streaming activity on certain of the AI developer’s content.
Needless to say, there’s ample incentive for profit-minded individuals to pump out songs via Boomy (as well as different AIs) and then attempt to bolster their stream counts in order to collect larger royalty payments. According once again to Boomy’s website, users will receive 80 percent of their projects’ streaming royalties “net of distribution fees.”
Boomy Pro subscribers can release an unlimited number of tracks, and Boomy’s offering likewise includes options for auto-generated song titles, artist names, and cover art. Consequently, with no real investment (financially, creatively, or in terms of time) at risk, it’s unclear whether Spotify’s bans will dissuade these same enterprising persons from taking similar steps moving forward.
To be sure, more than a few observers have pointed out the multitude of “artist” profiles through which seemingly AI-generated “songs” are pouring onto streaming services. In general, said profiles have random names and generic photos, while many of their works seem to bear substantial similarities in sound, length, and more.
Longer term, it goes without saying that successfully flagging a miniscule portion of one company’s AI projects for alleged fake streams isn’t a viable solution to the overarching problems presented by an endless stream of machine-made music. Universal Music (which funded an AI music company in 2022) and Warner Music Group (WMG) say they’re collaborating with streaming services on reforms, though neither company has publicly named Spotify as a partner on the initiatives.
Beyond the proliferation of AI songs attributed to new artist profiles, artificial intelligence is powering works made to sound like releases from Drake, The Beatles, Kanye West, and an abundance of others. Additionally, Grimes has rolled out a platform that enables fans to create songs with an AI depiction of her voice in exchange for a piece of the resulting royalties.