Microsoft denies "each and every allegation" in the consolidated copyright infringement class action alleging Microsoft and OpenAI committed “systematic theft on a mass scale” to feed and train their AI algorithms (see 2402050037), said its answer Friday (docket 1:23-cv-08292) in U.S. District Court for Southern New York in Manhattan. The claims by 29 authors and the Authors Guild fail because any unauthorized copies of any registered copyrighted works “constitute fair use,” said Microsoft’s answer. The claims also fail because all the algorithms accused of infringement, and all Microsoft’s products, services or actions in connection with those algorithms, “have commercially significant noninfringing uses,” it said. The claims of infringement are also barred by the Digital Millennium Copyright Act's safe harbor provisions, it said. The plaintiffs seek “improper damages” in violation of the Constitution and other applicable law, it said. “Any award of statutory or enhanced damages would constitute an unconstitutional penalty under the circumstances of this case, and would violate the due process and equal protection guarantees, and other substantive and procedural safeguards” afforded by the Constitution, it said. OpenAI asserted similar defenses in its separate answer Friday to the consolidated complaint.

The owner of a Bryan, Texas, bar and grill denies that his establishment publicly performs musical compositions in BMI's repertoire without a license or authorization, said his brief answer Wednesday (docket 4:24-cv-00199) to BMI’s complaint in U.S. District Court for Southern Texas in Houston. Yesterdays owner Jason Seymour asks that the relief requested in the complaint is denied and that he “recover all costs and expenses incurred in the defense of this case,” said his answer. BMI and 19 music publishers sued Seymour and his establishment Jan. 18, alleging nine claims of willful copyright infringement (see [Ref:2401190003).

The 14 plaintiffs in the first-filed copyright infringement suit against OpenAI in the Northern District of California seek to intervene and to dismiss the four actions against OpenAI and Microsoft filed subsequently in the Southern District of New York, said their motion Monday (docket 1:24-cv-00084). In the alternative, they seek to stay the four SDNY actions or transfer them to the NDCA, it said. The NDCA complaint was the first in the U.S. to allege that OpenAI committed direct copyright infringement when it made copies of the plaintiffs’ books without permission in order to train OpenAI’s language models, it said. The SDNY lawsuits are all “copycat cases,” said the motion. The subsequently filed SDNY actions are “strikingly similar” to the first-filed case, it said. Their claims “share common or overlapping theories of liability,” it said. OpenAI, in an apparent attempt to "undercut" the NDCA’s scheduling order, agreed to forego its motion to transfer the SDNY cases to NDCA, it said. By its “blatant forum shopping,” OpenAI is creating the likelihood, or certainty, of “inconsistent rulings in overlapping class actions and the attendant waste of judicial resources,” it said. “This is precisely the type of procedural gamesmanship the first-to-file rule was adopted to arrest,” it said. To avoid duplicative efforts, judicial waste and potentially disparate rulings, the SDNY should apply the first-to-file rule and dismiss, or in the alternative, stay or transfer the SDNY actions, it said.

U.S. District Judge Sidney Stein for Southern New York in Manhattan signed an order Tuesday (dockets 1:23-cv-08292 and 1:23-cv-10211) appointing Lieff Cabraser, Susman Godfrey and Cowan DeBaets as the plaintiffs' interim co-lead counsel in the newly consolidated copyright infringement class action brought by 29 authors and the Authors Guild against Microsoft and OpenAI (see 2402050037). The authors allege that OpenAI and Microsoft copied their works and then fed them into their large language models, algorithms designed “to output human-seeming text responses to users’ prompts and queries.” Those algorithms are at the heart of the defendants’ “massive commercial enterprise,” and at the heart of these algorithms “is systematic theft on a mass scale,” alleges the complaint.

Plaintiff Aleksandar Kavcic and defendant Broadcom have reached agreement on the terms of a settlement “and are in the process of preparing a stipulation of dismissal in accordance with such terms,” said their joint notice of settlement Monday (docket 3:20-cv-01246) in U.S. District Court for Northern California in San Francisco. The parties intend to file the dismissal within seven days and ask that the bench trial set for Feb. 13-14 be vacated and taken off the calendar, said the notice. Kavcic sued Broadcom in February 2020 for a declaratory judgment affirming his right to participate in defending the validity of his patent for detector technology for disk drives. Carnegie Mellon University, the patent’s assignee, sued Broadcom for infringement in July 2018, and Broadcom responded by attacking the patent’s validity.

U.S. District Judge John Tharp for Northern Illinois in Chicago granted Hytera Communications’ motion to reset its criminal trial date to Sept. 30 on allegations that Hytera and its engineers stole Motorola's trade secrets, said the judge's docket entry notification Friday (docket 1:20-cr-00688). There will be “no further extensions absent emergency circumstances,” said his notification. Hytera’s lead counsel, Steptoe’s Rachel Cannon in Chicago, asked for the postponement in a Jan. 11 motion, citing the strains of preparing for and participating in a separate “multi-week, four-defendant criminal trial” she's involved in that's scheduled to begin July 8. The Hytera trial delay also gives Tharp the chance to “consider the timing of the pending ruling” by the 7th U.S. Circuit Court of Appeals in the civil case between Motorola and Hytera, said the motion. That ruling “could establish important precedent for this case and otherwise impact the trial’s conduct,” it said. Though no one knows when the 7th Circuit will rule in the civil appeal, waiting to begin the Hytera trial until Sept. 30 will “maximize the chances” the circuit will rule before the start of this trial, it said. Hytera’s motion to dismiss the government’s criminal indictment for failure to present evidence of trade secrets to the grand jury (see 2401220002) “is taken under advisement,” said Friday's notification. A grand jury in May 2021 returned an indictment listing multiple counts of trade secret theft against Hytera and seven of its engineers who developed digital mobile radios for Motorola in Malaysia beginning in 2004 (see 2301260060). The engineers quit Motorola in 2008 and 2009 to go to work for Hytera in Shenzhen, and the government alleges that they took Motorola’s DMR trade secrets with them when they left.

OpenAI and Microsoft have built a business valued into the tens of billions of dollars “by taking the combined works of humanity without permission,” said 29 authors, plus the Authors Guild, in a class action Friday (docket 1:23-cv-08292) in U.S. District Court for Southern New York. The action combines the two previously separate complaints brought by fiction and nonfiction authors (see 2309210022). Rather than pay for intellectual property, such as when a person buys a book, “they choose to operate as if the laws protecting copyright do not exist,” the complaint said. The plaintiffs seek redress for the defendants’ “flagrant and harmful infringements” of their registered copyrights, it said. OpenAI and Microsoft copied their works and then fed them into their large language models (LLMs), algorithms designed “to output human-seeming text responses to users’ prompts and queries,” it said. Those algorithms are at the heart of the defendants’ “massive commercial enterprise,” it said. At the heart of these algorithms “is systematic theft on a mass scale,” it said. In training their LLMs, OpenAI and Microsoft “reproduced copyrighted texts to exploit precisely what the Copyright Act was designed to protect: the elements of protectible expression within them, like the choice and order of words and sentences, syntax, flow, themes, and paragraph and story structure,” it said. OpenAI and Microsoft “copied and data-mined” the works of writers, without permission or compensation, to build a machine that’s capable, or soon will be capable, “of performing the same type of work for which these writers would be paid,” it said. Without the “wide corpus” of copyrighted material on which to feed, there would be no ChatGPT, it said. The defendants’ commercial success “was possible only because they copied and digested the protected, copyrightable expression contained in billions of pages of actual text, across millions of copyrighted works -- all without paying a penny to the authors of those works,” it said.

In light of settlement discussions that appear to be “ongoing” between BMI and Maggie’s Bar & Grill in Jackson, New Jersey, U.S. Magistrate Judge Tonianne Bongiovanni for New Jersey in Trenton ordered the parties to submit an additional status update regarding those settlement talks by March 6, said his text order Thursday (docket 3:23-cv-03135). BMI contacted the restaurant’s managers more than 80 times since October 2019 about the necessity of “purchasing a license for the public performance of musical compositions in the BMI repertoire,” said the June 7 complaint from BMI and five music publishers (see 2306080044). They allege six claims of willful copyright infringement against the establishment, based on its “unauthorized public performance of musical compositions.”

Substantial production of documents for all discovery requests served by Jan. 29 will be completed by June 14 in the two consolidated authors’ copyright infringement complaints against Microsoft and OpenAI, said a scheduling order (dockets 1:23-cv-08292 and 1:23-cv-10211) signed Wednesday by U.S. District Judge Sidney Stein for Southern New York in Manhattan. The plaintiffs had proposed a May 10 deadline and the defendants, Aug. 31 (see 2401280001). Fact discovery in the consolidated cases will be complete by Sept. 17 and expert discovery by Dec. 9, said the order. The plaintiff authors in the two lawsuits allege that Microsoft and OpenAI copied their protected works on a mass scale without consent or compensation, then fed those copyrighted works into their large language models to train their AI algorithms (see 2401120005).

Defendant Altice USA seeks to transfer the copyright infringement case brought by 54 record labels and music publishers to the Eastern District of New York where it’s headquartered, said its motion Monday (docket 2:23-cv-00576) in U.S. District Court for Eastern Texas in Marshall. The Eastern District of New York is a “clearly more convenient venue” than the Eastern District of Texas, said the motion. All relevant evidence is derived from New York or locations “more proximate” to New York, it said. Two New York-based music conglomerates, Warner and Sony, through their subsidiaries, 35 of which are based in New York, “have sued New York-based Altice based on corporate decisions made in New York,” it said. The bulk of the evidence is in New York, and most of the parties’ current and former employees are in that state, said the motion. The plaintiffs’ notices purportedly identifying alleged activity of Altice’s users were sent to, and processed by, Altice in New York, it said. The defendant’s decisions about what to do in response to the plaintiffs’ accusations “took place at its headquarters in New York,” it said. Altice’s data related to notices of alleged infringement and accused users are maintained in New York, and the plaintiffs’ corporate decisions and conduct took place in their respective New York headquarters, it said. The recording industry's Dec. 7 complaint alleges that Altice has knowingly contributed to, and reaped “substantial profits” from, massive copyright infringement committed by thousands of its internet subscribers (see 2312080050). Altice's motion to dismiss Monday contends that the recording industry in recent years has wrongly sought to “install” ISPs as the internet’s “copyright police" (see 2401300001).