The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Sharp Words from Judge in Google AI Copyright Case
From today's Order Amending Case Schedule by Judge Eumi Lee (N.D. Cal.) in In re Google Generative AI Copyright Litigation:
The parties in this case have helped themselves to more judicial resources than necessary, given the experienced counsel representing each side. Lawyers should vigorously advocate on their client's behalf. But lawyers should also be problem-solvers. Unfortunately, a good deal of the problem-solving work in this case has been shifted to the Court. The latest example of this trend is the parties' inability to compromise on a reasonable extension of the case schedule. See Mot. to Extend, ECF No. 173. The Court did not need six briefs to decide whether to amend the case schedule, but that is what the Court received. See ECF Nos. 173, 182, 189, 193, 195-1, 196. Having reviewed all these papers, the Court GRANTS the motion in part, albeit reluctantly.
"A schedule may be modified only for good cause and with the judge's consent." The good cause inquiry "primarily considers the diligence of the party seeking the amendment." "Good cause will be shown if a scheduling order's deadline 'cannot reasonably be met despite the diligence of the party seeking the extension.'" But "[i]f that party was not diligent, the inquiry should end."
The primary motivation for Plaintiffs' motion to amend the case schedule is the fast- approaching August 8 deadline to move for class certification, which Plaintiffs wish to extend by three months. In many respects, Plaintiffs have not acted diligently to meet this deadline. At the December 18, 2024 case management conference, the Court instructed Plaintiffs to focus their initial discovery efforts on information that is necessary for class certification. The Court set the August deadline as a compromise between the approaches proposed by the parties. If Plaintiffs had focused their efforts, there was enough time to complete class certification discovery in the past eight months. {The Court notes that the plaintiffs in Bartz v. Anthropic PBC, No. 3:24-cv-05417-WHA, moved for class certification less than six months after the initial case management conference.}
Unfortunately, Plaintiffs have wasted time on distractions. First, Plaintiffs improperly amended the class definition to create a fail-safe class, triggering avoidable litigation. In its April 21 order on that issue, the Court warned Plaintiffs that it would "not find good cause for an extension [of class certification deadlines] if the record reflects Plaintiffs' failure to prioritize and diligently advance discovery on issues relevant to class certification." Second, since April, Plaintiffs have wasted time seeking overbroad discovery, failing to prioritize issues relevant to class certification. Remarkably, Plaintiffs now say they would use an extension in part to "negotiate or litigate additional [discovery] disputes"—including disputes over "privilege claims, 30(b)(1) witnesses, production, [and] corporate witness preparedness." More discovery litigation is not good cause for an extension of class certification deadlines.
However, the Court finds good cause for a 45-day extension of Plaintiffs' deadline to move for class certification. This extension provides Plaintiffs additional time to review discovery that Google has produced in the past six weeks, including training data and custodial productions. It appears that Plaintiffs first accessed Google's training data for the generative AI models at issue in early July, and Plaintiffs report some lingering issues. Review of the training data, at least, is plausibly relevant to class certification, and the current schedule does not afford Plaintiffs sufficient time to review that material, given its recent production.
However, the Court does not accept Plaintiffs' representation that a three-month extension is needed. Plaintiffs' request for three months includes time to litigate discovery issues unrelated to class certification, which the Court does not approve. And Plaintiffs could have used the time available to prioritize training data, instead of engaging in discovery skirmishes on ancillary issues. The amended class certification schedule is reflected below.
It is unclear why compromise seems to be so difficult in this case. One potential issue is that Plaintiffs have lacked focus, which is essential to a putative class action of this magnitude. Another potential issue is that Google has been somewhat intransigent in discovery, perhaps due to its view that a class will never be certified.
The Court welcomes creative proposals as to how the Court might align incentives and encourage compromise going forward. The Court is not keen to issue a middle-ground ruling on every dispute presented, which would only incentivize both sides to take extreme positions from the outset.
The Court would prefer not to issue orders like this one. The parties have been told several times that they need to make greater efforts to compromise, yet disputes persist and proliferate. Access to the courts is a public good, and time spent on issues like this one detracts from other important work—including other work in this case.
For the foregoing reasons, the following unexpired deadlines in the Case Management and Scheduling Order, ECF No. 88, are AMENDED as shown below. By August 13, 2025, the parties shall submit a proposed order to amend the other case deadlines through trial, consistent with the amended class certification deadlines below. This is not an invitation to relitigate the general structure of the case schedule or to substantially extend the close of fact discovery. The Court will not consider competing proposals….
Show Comments (4)