In a strategic move to preserve their right to seek reconsideration of previously dismissed DMCA § 1202(b) claims, the plaintiffs in Andersen v. Stability AI have voluntarily dismissed with prejudice all DMCA claims. The opportunity to seek reconsideration of the dismissed claims will come if a reversal occurs in the Doe 1 v. Github interlocutory appeal. We covered more of the decisions in the Github case in this prior post.

The Court in Andersen recently dismissed the plaintiffs DMCA § 1202(b) claims when it adopted the Github court’s “identicality” requirement for DMCA § 1202(b) claims. This requirement posits that for a DMCA copyright management information (CMI) removal claim to be valid, there must be “identicality” between the original work and the copy from which CMI has been removed. Other district courts addressing similar issues have reached a different result.[1] The Ninth Circuit Court of Appeals will address the issue in the Github interlocutory appeal.

The Andersen court’s order dismissing the plaintiff’s DMCA § 1202(b) claims also granted plaintiffs leave to file a Second Amended Complaint (SAC). Compliance with the court order required filing the SAC without the DMCA plaintiffs. As this could potentially compromise the plaintiff’s rights to seek reconsideration, the plaintiffs initially included the DMCA plaintiffs in the SAC. Seeking to obey court order and retain reconsideration rights, the plaintiffs voluntarily dismissed with prejudice all DMCA claims in exchange for the defendants agreeing not to challenge future reconsideration of the DMCA claims based on their omission from the SAC.

The Ninth Circuit’s pending decision will reach far beyond the Github and Andersen cases. Should the Ninth Circuit affirm the Github decision, AI companies will have some level of comfort with respect to CMI claims, at least in the Ninth Circuit, as the outputs of AI models are generally not identical copies of the data they are trained on. If the Ninth Circuit reverses Github, plenty more litigation about CMI will likely ensue to address when the “substantial similarity” of AI model outputs to their training data will require maintaining CMI.

FOOTNOTES

[1] In Fischer v. Forrest, the Southern District of New York found that a Section 1202(b) claim could proceed without requiring the removed or altered CMI to be from an identical copy of the work. The court emphasized the intent behind the removal or alteration of CMI rather than the exactness of the copy. In GC2 Inc. v. Int’l Game Tech. PLC, the Northern District of Illinois held that the removal or alteration of CMI did not need to be from an identical copy of the work. The court focused on whether the removal or alteration facilitated infringement, regardless of whether the work was an exact copy