OpenAI’s Motion to Dismiss Reads ‘More Like Spin Than a Legal Brief,’ Says N.Y. Times

Lacking any “real grounds” for dismissal, OpenAI devotes much of its Feb. 26 motion to dismiss the New York Times’ copyright infringement complaint (see 2312270044) “to grandstanding about issues on which it hasn’t moved,” said the Times’ memorandum of law Monday (docket 1:23-cv-11195) in opposition. OpenAI’s motion introduces no fewer than 19 “extrinsic documents,” none of which can be properly considered in a motion to dismiss, “in a submission that for nearly 10 pages reads more like spin than a legal brief,” it said. OpenAI’s “attention-grabbing claim” that the Times hacked its products “is as irrelevant as it is false,” it said. As the complaint makes clear, the Times “elicited examples of memorization by prompting GPT-4 with the first few words or sentences of Times articles,” it said. That work was necessary only because OpenAI doesn’t disclose the content it uses “to train its models and power its user-facing products,” it said. Yet in OpenAI’s telling, the Times engaged in wrongdoing by detecting OpenAI’s theft of the Times’ own copyrighted content, it said. OpenAI’s “true grievance” is not about how the Times conducted its investigation, “but instead what that investigation exposed,” said the memorandum. It found that OpenAI and Microsoft built their products by copying the Times’ content “on an unprecedented scale,” it said. OpenAI doesn’t and can’t dispute that fact, the memorandum said. Despite seeking to justify this conduct however it can, OpenAI doesn’t move to dismiss the lead claim that it infringed the Times’ copyrights to train and operate its latest models, it said. Against those claims that it challenges, OpenAI advances “mainly factual arguments that cannot be decided on the pleadings,” it said. To support its statute-of-limitations argument for claims based on models developed before December 2020, OpenAI asks the court to make a factual finding that the makeup of the datasets used to train those models was common knowledge in 2020, it said. OpenAI’s bid to dismiss the contributory infringement claim “turns on disputed facts about user behavior,” it said. OpenAI’s attack on the Digital Millennium Copyright Act claim turns on disputed fact issues concerning the design of its model-training process, which can’t be resolved before discovery into that design, said the memorandum. OpenAI further asks the court to dismiss the unfair competition by misappropriation claim by ignoring the Times’ allegations of OpenAI and Microsoft “free-riding” and deciding that the Times’ product recommendations aren’t generated by efforts akin to reporting, said the memorandum. That’s “yet another premature argument,” it said: “Discovery, not dismissal, is warranted to resolve each of these well-pleaded claims.”