US Judge Dismisses Beaulier v Roblox NoAI AI Training Lawsuit

• A federal judge dismissed a putative class action against Roblox concerning the alleged use of an artist's 3D models for AI training.
• Artist Austin Beaulier claimed Roblox removed his "NoAI" tags, which he argued constituted protected content management information under the DMCA.
• U.S. District Judge Beth Labson Freeman, a Barack Obama appointee, ruled that while Beaulier adequately pleaded standing and injury, he failed to sufficiently demonstrate Roblox's actions were intentional.
• The judge stated that merely showing CMI removal as a side effect of a uniform process is insufficient to prove intentionality under DMCA Section 1202.
• Following an initial dismissal with leave to amend, U.S. District Judge Beth Labson Freeman has issued a second dismissal of Beaulier's CMI claims against Roblox, again for failure to adequately allege intentional removal.

Pleadings that merely establish that a defendant employed a uniform process or a process in which removal is a side effect are insufficient to plead intentional removal and fail to 'nudge' allegations of intent 'across the line from conceivable to plausible.'

A federal judge in San Jose, California, has dismissed a putative class action brought against the online gaming platform Roblox, which alleged the company used an artist's 3D models to train its artificial intelligence systems without permission. The lawsuit, filed by 3D model artist Austin Beaulier in March, claimed that Roblox utilized his works despite them being marked with a specific "NoAI" tag, intended to prohibit such use. Beaulier asserted that these tags were removed during Roblox's processing of his content for AI training purposes.

The artist's complaint detailed that Roblox employed various 3D model datasets, including one named "Obiaverse-XL," which reportedly contained millions of 3D works sourced from publicly accessible repositories. Beaulier discovered his own creations within this dataset. The lawsuit contended that Roblox's process for training its generative AI models involved converting raw creative works, such as 3D models, into numerical inputs. This conversion allegedly occurred through "machine-learning preprocessing pipelines" designed to transform raw 3D assets into suitable training inputs for generative AI systems, a process during which the "NoAI" tags were purportedly removed.

Central to Austin Beaulier's case was the argument that the "NoAI" tag constituted "content management information" (CMI) under the Digital Millennium Copyright Act (DMCA). The DMCA defines CMI to encompass terms and conditions governing the use of a work, making its intentional removal a violation of federal law. Beaulier accused Roblox of intentionally removing these tags while preparing his works for its AI training pipelines.

U.S. District Judge Beth Labson Freeman, a Barack Obama appointee, presided over the case. In her 15-page ruling, Judge Freeman acknowledged that Beaulier had sufficiently pleaded both standing and injury. However, the critical flaw in the artist's argument, according to the court, was his failure to adequately demonstrate that Roblox's actions were intentional, a key requirement for a claim under DMCA Section 1202 intentional removal.

Judge Freeman's decision underscored the significant burden on plaintiffs to prove intentionality in cases involving the alleged removal of content management information. The court found that Beaulier's factual allegations amounted to little more than "conclusory assertions" regarding the presence of CMI before and its absence after Roblox's preprocessing pipeline. This lack of specific evidence regarding intent proved fatal to the initial complaint.

In her ruling, Judge Freeman explicitly stated that pleadings which merely establish a defendant's use of a uniform process, or a process where the removal of CMI is an incidental side effect, are insufficient to plead intentional removal. She emphasized that such allegations fail to "nudge" the claim of intent "across the line from conceivable to plausible." For a claim to proceed, Beaulier "must plead something more" than the mere outcome of a process, requiring concrete evidence of deliberate action by Roblox to remove the "NoAI" tag.

Following the initial dismissal with leave to amend, Austin Beaulier filed an updated complaint. However, U.S. District Judge Beth Labson Freeman has since issued a second dismissal of Beaulier's CMI claims against Roblox, again for failure to adequately allege intentional removal of CMI. This second dismissal reinforces the high bar for proving intentionality under DMCA Section 1202 and provides further guidance for future cases involving generative AI copyright infringement.

This outcome in the Beaulier v Roblox NoAI dismissal case highlights the challenges artists face in proving deliberate infringement when their works are incorporated into large-scale AI training datasets. Attorneys representing both the plaintiff and Roblox did not immediately respond to requests for comment following the ruling.

This ruling clarifies the high bar for proving intentional removal of Content Management Information under DMCA Section 1202 in AI training contexts. Lawyers advising clients on AI data usage or intellectual property protection must understand this precedent, emphasizing the need for robust evidence of intent in such claims.

Source: Original reporting via Courthouse News Service