Twenty briefs, published by the defendant
OpenAI published excerpts from twenty amicus briefs on 30 October, all of them filed in support of its fair use defence in the consolidated copyright litigation in the Southern District of New York. That proceeding includes The New York Times Co. v Microsoft and OpenAI, and the post describes the newspaper’s suit as a “narrow pursuit of an undeserved payday”.
The list is not what an AI company’s supporters usually look like. Alongside the trade bodies — the Computer and Communications Industry Association, Engine and the Developers Alliance, the Business Software Alliance — are the Electronic Frontier Foundation, Public Knowledge, the American Library Association, the Association of Research Libraries and the American Council of the Blind. The venture firm Khosla Ventures filed. So did the law professors Matthew Sag, Ed Lee and Rebecca Tushnet, the economist Joshua S. Gans, the writer Kevin Kelly and the Pulitzer-winning journalist Ron Suskind.
The government’s brief
The entry OpenAI lists last is the United States Government. Its excerpt argues that the Times “seeks to narrow fair-use doctrine to exclude the training of OpenAI’s large language models”, and that doing so would be “inconsistent with basic copyright law principles”. The brief closes on economics rather than law: “Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility.”
The Justice Department first intervened on the AI companies’ side on 1 September. Three sitting House Democrats — Jerry Nadler, Jamie Raskin and Hank Johnson — filed against that position on 30 October, arguing the Copyright Act contains no national security exception.
Two former members of Congress filed the other way. Senator Rick Santorum and Representative Rick Boucher told the court that “the freedom of American scientists, companies, and entrepreneurs to train AI is of paramount economic and strategic importance”. The legislature is now on both sides of the brief, depending on whether its members still sit.
The narrower fight over dilution
Several briefs converge on one question rather than the broad one. The Authors Alliance asks the court to hold that “market dilution” — aggregate competitive pressure from outputs that share a genre or subject matter without copying protected expression — is not cognisable market harm under 17 U.S.C. § 107(4). Ed Lee and Rebecca Tushnet put the same point more bluntly: “Genres and facts are uncopyrightable.”
That is the theory the Times needs. If dilution counts as harm under the fourth fair use factor, training can lose on that factor without any output resembling an article.
The other side filed the same day
A 31-page brief for the Times was filed by the Human Consent Foundation, the non-profit led by Cate Blanchett, joined by the talent agency CAA, the Music Artists Coalition, the National Association of Voice Actors and the Creators Coalition on Artificial Intelligence, according to TheWrap, which obtained the document. An appendix lists Anne Hathaway, Bette Midler, Emma Thompson, Helen Mirren, Dan Levy, Ron Howard, Madonna and Olivia Wilde as supporters.
Its argument is practical rather than doctrinal: the defendants say clearing rights at scale is prohibitively hard, and the creative industries already do it. The brief points to collective licensing through CISAC and Broadcast Music, Inc., and to chain-of-title and digital rights management practice. Blanchett is quoted as saying: “If someone wants to use my face, my voice or my image, they ask me first.” It followed an earlier brief from SAG-AFTRA, the Writers Guild of America and the American Federation of Musicians.
What to watch
None of the filers has posted the briefs themselves, and the amicus submissions are on the public docket rather than on the parties’ sites, so the excerpts OpenAI chose are the only published version of most of them. Judge Sidney Stein has not said which briefs he will accept. No appellate court has yet ruled on whether training a model on copyrighted work is fair use.
