Twitter Can’t Quash a 512(h) Subpoena

This case involves the Twitter account “@CallMeMoneyBags,” a pseudonymous account. In October 2020, the user posted 6 tweets attacking Brian Sheth, a private equity billionaire, that included photos “depicting a woman or portions of a woman’s body.” The opinion doesn’t clarify who is depicted in the photos or who owns those photos, other than to note that “Bayside registered its copyrights in the photos on November 2, 2020.” The relationship between Sheth and Bayside is unclear; but allegedly Bayside isn’t owned or controlled by Sheth, leaving open the important and obvious question of why Bayside is undertaking this enforcement action. After some back-and-forth, Bayside submitted a takedown request through Twitter’s online form and Twitter honored it.

Bayside followed up with a 512(h) subpoena request. If you don’t recall this procedure, it says that after sending a 512(c)(3) takedown notice, a copyright owner can get, as a matter of right, an unmasking subpoena. Back in 1998, Congress didn’t appreciate how unmasking subpoenas could be weaponized, but we know better now, which makes 512(h)’s easy access to unmasking subpoenas a dangerous relic of a different Internet era. As required by law, the court clerk issued Bayside’s requested subpoena, and Bayside served it on Twitter.

(More background on this odd case from Forbes and Techdirt).

Twitter moved to quash the subpoena, essentially trying to require the court to conduct the typical First Amendment review that applies to non-512(h) unmasking subpoenas. The court says that instead of doing a First Amendment analysis, it’s possible that a fair use analysis is sufficient for 512(h) subpoenas (citing Eldred for the principle that fair use is the First Amendment safety valve to copyright infringement). The court says:

if the fair use inquiry demonstrates that an anonymous speaker is not an infringer of copyrighted works, a DMCA subpoena must be quashed, since the only authorized purpose for a subpoena under the DMCA is to discover the identity of an alleged infringer

However, @CallMeMoneyBags no-showed, despite dubious attempts at service. This leaves some evidentiary gaps in the fair use analysis:

The tweets at issue consist of statements combined with Bayside’s photos along with hashtags that do not convey an obvious meaning. Without evidence regarding the purpose and meaning of the tweets, the court cannot say that @CallMeMoneyBags’s use of the photos was “transformative” or reasonable in relation to @CallMeMoneyBag’s purpose in posting the tweets…

the absence of any evidence from the user(s) of @CallMeMoneyBags about the relevant market or the lack of market harm from the publication of Bayside’s photos precludes a finding of fair use here

Due to the failed fair use defense, the court orders Twitter to comply with the 512(h) subpoena.

I’m still scratching my head wondering why Twitter chose to double-down on this particular case. Normally, if Twitter wanted to establish favorable legal precedent on a topic like 512(h), it would pick sympathetic facts and a cooperative user. Twitter seemed to extend itself here without the necessary ingredients for success. So yay for Twitter standing up against potentially overreaching 512(h) subpoenas, and I hope it will try again when it has a better chance of winning.

[This is my last blog post for 2021. Thanks for reading the blog this year! Wishing you and your family have a wonderful 2022.]

Case citation: In re DMCA § 512(h) Subpoena to Twitter, Inc., 2021 WL 6135300 (N.D. Cal. Dec. 29, 2021)

Other Blog Posts on 512(h)