This newest subpoena showdown, having been heating up behind the scenes throughout 2026, quietly boiled over last week. That’s when Suno fired off a motion to compel the Mechanical Licensing Collective’s (MLC’s) compliance with a seemingly straightforward documents request.
With the majors’ royalty- and market-dilution allegations front of mind, the AI platform in April set its sights on a variety of label-specific consumption figures dating back to 2021.
Letting the actual subpoena take the wheel for a moment, Suno sought: “For Amazon, Apple, Deezer, Meta, Pandora, Snapchat, SoundCloud, Spotify, Tidal, TikTok, and YouTube, respectively, Documents sufficient to show performance data for all Labels, disaggregated on a Label-by-Label basis, in Your Database on a monthly basis for the period of 2021 to present.”
(Among different things, “performance data” includes ISRC, title, label, distributor, play count, use type, match status, and featured acts. Also, per the MLC’s opposition, the entity “does not collect data directly from record labels or sound recording distributors, does not maintain data on behalf of record labels or sound recording distributors, and does not have mandatory reporting fields for record label or sound recording distributor information.”)
In the platform’s own words, this sizable collection “of music performance data provides a comprehensive record of what happened in the music streaming market during the key period when Suno’s platform was launched into the market.”
Furthermore, said “data is highly relevant to what actually occurred before and after AI-generated music arrived and is therefore directly relevant to Plaintiffs’ central theory that AI-generated music caused harm to their market share, royalties, and so on,” according to the motion.
“For Amazon, Apple, Deezer, Pandora, SoundCloud, Spotify, Tidal and YouTube, respectively, Documents sufficient to show Performance Data for all Labels in Your Database on a monthly basis for the period 2021 to present,” the dialed-back and clarified August subpoena reads.
If left at that, the episode, complete with a laundry list of objections from the MLC concerning the “vague, ambiguous, and overbroad” ask, could safely be described as another testament to the increasingly aggressive AI litigation landscape, where subpoenas are reaching a growing number of non-parties.
However, there’s a bit more to the situation. The way Suno’s counsel tells the story, multiple months, emails, phone calls, and billable hours later, the MLC committed to providing several terabytes’ worth of data – with the AI platform agreeing to foot the roughly $5,000 bill – before abruptly reversing course at the behest of DSPs.
But we needn’t retread the voluminous correspondence here; after the second subpoena surfaced, the Pryor Cashman-repped MLC opposed the request in part by stating that it’d “received objections to disclosure from affected” DSPs.
Fast forward to September 23rd, when The Campbell Firm’s Chris Campbell, who’s repping both Suno and Udio, offered a summary of his own in a letter to the appropriate Pryor Cashman attorneys.
“We have spent nearly four months working with the MLC to focus this production on relevant material and to eliminate any burden, including by agreeing to pay the MLC’s costs,” he vented. “The MLC confirmed it could produce the data, provided a cost estimate, and advised the DSPs that it intended to produce in short order—and then suddenly reversed course and refused.”
With that, it’ll be worth tracking the heated subpoena clash moving forward; at present, it doesn’t appear that a dataset compromise (meaning tailoring the disclosure to the plaintiff majors’ data) will do the trick, either.
“You also asked whether Suno and Udio would accept data only for the Plaintiff labels, but not any other labels included in the MLC’s database. We have considered it, and a limitation of that kind would defeat a central purpose of the request,” Campbell wrote.
And in the bigger picture, the growing list of non-party subpoenas – and copyright lawsuits – involving AI developers doesn’t exactly suggest that a wave of breakthrough settlements will materialize anytime soon.
