The Stability AI defendants have filed their opposition to Sarah Andersen’s request to Judge Orrick to reverse Magistrate Judge Cisneros’s discovery ruling barring the plaintiffs’ expert Dr. Ben Zhao from accessing the confidential datasets and information of the defendant AI companies.
Zhao is famous for developing various tools–including Nightshade–that “poison” images to “disrupt” the effectiveness of image diffusion models. Indeed, Dr. Ben Zhao’s webpage at University of Chicago has the title: “Ben Zhao: Nightshade: Data Poisoning to Fight Generative AI.”
Stability AI asserts: “While this dispute has been pending, Dr. Zhao co-authored a new paper on his latest data poisoning tool, ‘Hemlock,’ and announced plans for its public release ‘for potential use by visual artists.’” Zhao et al., “On the Feasibility of Poisoning Text-to-Image AI Models via Adversarial Mislabeling” at 12 (June 27, 2025), https://arxiv.org/pdf/2506.21874 (Gratz Decl. Ex. A).
In her prior ruling, Judge Cisneros concluded:
The Court disagrees with Plaintiffs’ argument that Dr. Zhao is not a competitor because he is an academic researcher rather than part of a company that directly competes “for the same dollars” as Defendants. Id. at 3 n.1. His work is “in functional competition with Defendants” as he develops tools that attack Defendants’ generative AI models. Id. at 5; see Voice Domain Techs., LLC v. Apple, Civil Action No. 13-40138, 2014 WL 5106413, at *4 (D. Mass. Oct. 8, 2014) (“Even where parties are not traditional competitors in the market place, an individual may
still be deemed a [competitor] where the parties are in an adversarial posture and the individual receiving the highly confidential information would be especially situated to take position that are directly harmful and antagonistic to the defendant.”) (citation modified). Dr. Zhao attests that his tools teach generative AI models to make mistakes: a model trained on enough Glaze- or Nightshade-modified images might produce an image in the style of Jackson Pollock when prompted to produce a realistic charcoal portrait or produce an image of a leather purse when prompted to draw a cow. Zhao Decl. ¶¶ 10-11. As Dr. Zhao researches and creates tools that make generative AI models such as Defendants’ less reliable, his work impacts “Defendants’ ability to operate in the marketplace.” 2 ECF No. 300 at 5; see Tomahawk Manufacturing, 344 F.R.D. at 471-72 (identifying that “special concerns arise when prospective experts or consultants may themselves be competitive with the disclosing party’s business”).
Related Posts