(By You Yunting) The New York Times v. OpenAI and Microsoft—a landmark case in artificial intelligence—has seen one more dramatic twist. In the pretrial proceedings in the consolidated cases, the defendants accused The New York Times of laundering evidence: a paper cited by the plaintiff’s expert was allegedly funded by The Times’ own law firm and the funding was never disclosed to the court. The story has drawn renewed media attention. Today, we will examine whether this evidence-laundering will materially affect the lawsuit — and what the consequences would be if similar conduct occurred in China.
What exactly is “evidence laundering”? In short, the law firm representing The Times funded a research project; its expert witness then drew heavily on the resulting paper extensively in his testimony. OpenAI and Microsoft contend that the research was commissioned for purposes of the litigation, and that The Times deliberately concealed the funding relationship and the resulting conflict of interest in its disclosures and in the expert’s credentials, thereby presenting the evidence as cleaner and more objective than it actually was.
In December 2025, Susman Godfrey, the law firm representing The Times, provided $100,000 to fund a paper by Professor Tuhin Chakrabarty of Stony Brook University (State University of New York) on how AI-generative books dilute the market for human authors. However, when The Times identified the professor and his co-authors as potential experts, the résumés it submitted omitted any mention of the law firm’s funding. Later in the litigation, Lasinski, an expert retained by The Times, submitted a supplemental report to the court at the final stage that drew heavily on that very paper.
The paper ran full-text AI detection across 14,419 self-published genre-fiction titles on the Amazon US marketplace from 2023 to 2026, matched the results against daily sales data, and concluded that an influx of AI-generated books onto the platform crowds out and dilutes the sales and market returns of books by human authors. The paper argues that AI inflicts actual market harm on rights holders. If the court accepted, it could undermine OpenAI’s core defense: that using copyrighted content to train AI constitutes fair use. The defendants therefore moved to strike Lasinski’s supplemental report and to exclude the paper.
Not being a U.S.-qualified lawyer, I cannot judge how this will affect the case. Under Chinese law, however, the matter remains far from certain. China’s Civil Procedure Law provides a mechanism comparable to the U.S. expert witness system: a party may apply to the court for a person with specialized knowledge to appear in court and give opinions on specialized issues. The law does not prohibit law firms from funding academic research. However, when funded research is cited by an expert in an opinion submitted to the court, the funding behind it must be disclosed. Otherwise, the opinion risks misleading the court and may violate the good-faith principle in the Civil Procedure Law.
In this case, the written opinion of The Times’ expert would certainly have been reviewed by its counsel. Yet the expert report submitted to the court did not disclose the funding behind the cited paper. Whether deliberate or negligent, this constitutes a failure to disclose a conflict of interest that was required to be disclosed. Moreover, The Times’ expert did not cite the paper in his initial opinion; he introduced it only in a supplemental report filed at the final stage—rather like presenting surprise evidence at the end of the evidentiary period. Any observer can see the intent: to keep the facts hidden and slip the evidence past the court. Therefore, if this case were heard in China, The Times’ side would, at the very least, be open to criticism for a good-faith deficiency.
Even so, whether the report’s arguments are accepted turns on their own merit. The judicial interpretation of the Civil Procedure Law provides that opinions given in court by persons with specialized knowledge on specialized issues are deemed statements of the party concerned. Under the Supreme People’s Court’s Provisions on Evidence in Civil Proceedings, a party’s own statements cannot by themselves serve as the basis for establishing the facts of a case; the court must make a comprehensive determination based on the facts and evidence in the case. That said, by the court’s standards for evaluating evidence, whether a witness or evidence provider has an interested relationship with a party is also one of the factors the adjudicators weigh in reviewing evidence.
In my view, the citation of that very paper is not aimed at the court’s factual findings; rather, it seeks to influence the legal analysis by injecting a slanted viewpoint. From the perspective of the judge handling the case, the plaintiff’s conduct is indeed somewhat misleading. Even so, expert opinions submitted by both sides will inevitably favor the party who retained the expert, so the court would therefore scrutinize whether the opinion’s reasoning is sound on its own terms, and the influence on the court of a paper cited within an expert opinion is more indirect still.
Unless OpenAI can show that the funded paper contains fabricated data, this episode, although misleading, is not particularly serious in nature. The point OpenAI has seized on against The Times may therefore not be as serious as OpenAI’s own conduct—the concealment of key evidence discussed in my article “OpenAI Accused of Concealing Key Evidence: Is the Fair Use Defense in the Times Case in Jeopardy?” What OpenAI did involved bad faith as to facts of the case, which is far more serious than a mere failure to disclose that a cited paper was funded and its conclusions shaped by that funding.
Disclaimer of Bridge IP Law Commentary
Short Link: