20页纸，裁不了任何案子，却想影响所有案子。
9月2日，特朗普政府向纽约南区法院提交了一份20页的法庭陈述（brief），公开支持OpenAI的核心立场：未经授权使用版权材料训练大语言模型，属于合理使用。这起案件由《纽约时报》于2023年12月提起——如今，美国行政分支亲自下场，站到了原告对面。
引用的每一句都有出处
brief引用的背书来自特朗普2025年1月签署的行政令《Removing Barriers to American Leadership in Artificial Intelligence》。核心论点写得直白：美国有"强烈利益"去继续发展"强劲且有竞争力的人工智能产业"，并"在全球AI使用的实践与规则上树立标准"；另一句更关键——"在对合理使用原则的误解下去约束LLM开发，将扼杀这样的创造与科学进步，同时损害美国的繁荣与经济流动性"。翻译成大白话：版权问题，不许挡AI领导权的路。
分层看：行政站队，司法先例
这份文件的性质必须先说清：它是法庭陈述，不是裁决——写它的人没有管辖权。但行政分支站队的分量在于"风向"：它向法庭、也向所有同类诉讼的当事人，亮明了本届政府的偏好。第二层看司法：Alsup法官此前裁定的Anthropic 15亿美元和解，处罚的是盗版shadow library，而非训练行为本身；迄今为止，AI训练版权的判例总体偏向AI公司。行政偏好叠加司法走向，两条线目前指向同一个方向。
案子里另一根刺，别忽略
不过这场诉讼不只关于法理：2026年7月，《纽约时报》指控OpenAI在诉讼过程中隐藏证据——若这一指控在后续程序中被坐实，再多政府声援也未必扭转具体案件的事实认定。真正值得盯的还有域外效应：纽约南区法院是全球AI版权诉讼的参照系，这里的每一次风向变化，都会被其他法域的律师写进代理意见。一份brief改变不了判决，但它已经改变了辩论的起点。
明天见。
Twenty pages that decide no case — yet aim to influence all of them.
On September 2, the Trump administration filed a 20-page court brief in the Southern District of New York, openly backing OpenAI's core position: training large language models on copyrighted material without authorization constitutes fair use. The case was brought by The New York Times in December 2023 — now the executive branch itself has stepped in, on the opposite side of the plaintiff.
Every citation has a source
The brief's backing comes from the January 2025 executive order, "Removing Barriers to American Leadership in Artificial Intelligence." The core argument is blunt: the United States has a "strong interest" in continuing to develop a "robust and competitive artificial intelligence industry" that "sets the standard for the practice and procedure of AI use globally." The sharper line: "Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility." In plain terms: copyright shall not stand in the way of AI leadership.
Read it in layers: executive alignment, judicial precedent
First, the nature of the document: it is a court brief, not a ruling — its authors hold no jurisdiction. But the weight of the executive branch's alignment lies in the weather it signals: to the court, and to every party in similar litigation, it states this administration's preference. The second layer is judicial: Judge Alsup's approved $1.5 billion Anthropic settlement punished a pirated shadow library, not training itself; to date, AI-training copyright outcomes have broadly favored AI companies. Executive preference plus judicial drift — both lines currently point the same way.
Don't miss the other thorn in this case
Yet this lawsuit is not only about doctrine: in July 2026, The New York Times accused OpenAI of concealing evidence during the litigation — if that accusation is substantiated downstream, no amount of government support guarantees the fact-finding. The extraterritorial effect deserves watching too: the Southern District of New York is the reference frame for AI copyright litigation worldwide, and every shift in its weather gets quoted in briefs across jurisdictions. A brief changes no verdict — but it has already moved the starting point of the argument.
See you tomorrow.
一份brief改变不了判决，但它已经改变了辩论的起点。
—— Dawn Vision编辑部
A brief changes no verdict — but it has already moved the starting point of the argument.
— The Dawn Vision Editorial Desk
9月2日特朗普政府向纽约南区法院提交20页法庭陈述，支持OpenAI在NYT诉OpenAI案（2023年12月提起）中'未经授权使用版权材料训练LLM'的立场 · 引用特朗普2025年1月行政令《Removing Barriers to American Leadership in Artificial Intelligence》 · 原文论点：'strong interest in continuing to develop a robust and competitive artificial intelligence industry that sets the standard for the practice and procedure of AI use globally'；'Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility' · 性质：法庭陈述非裁决，brief作者无管辖权 · NYT曾于2026年7月指控OpenAI在诉讼中隐藏证据 · 先例：Alsup法官裁定Anthropic $1.5B和解（处罚盗版shadow library而非训练本身）；AI训练版权判例总体偏向AI公司
On Sept 2 the Trump administration filed a 20-page brief in the Southern District of New York backing OpenAI's position on training LLMs on copyrighted material in NYT v. OpenAI (filed Dec 2023) · Cites Trump's January 2025 executive order 'Removing Barriers to American Leadership in Artificial Intelligence' · Original arguments: 'strong interest in continuing to develop a robust and competitive artificial intelligence industry that sets the standard for the practice and procedure of AI use globally'; 'Constraining LLM development under a misunderstanding of fair use doctrine would thwart such creative and scientific progress while hindering American prosperity and economic mobility' · Nature: a court brief, not a ruling; its authors hold no jurisdiction · NYT accused OpenAI of concealing evidence in July 2026 · Precedent: Judge Alsup's approved $1.5B Anthropic settlement (punishing a pirated shadow library, not training itself); AI-training copyright outcomes broadly favor AI companies
Sources · 信源 Sources
本文基于 Dawn Vision 认知引擎处理当日采集信号并交叉验证 2 个核心来源后生成，经编辑部人工审核。素材来源：TechCrunch（brief提交与核心论点）、The Verge（诉讼背景与时间线）；brief原文论点以法庭文件表述为准。正文明确区分'法庭陈述'与'裁决'两种性质，未将政府立场表述为司法结论。
This article was generated by the Dawn Vision cognitive engine processing collected daily signals with cross-validation of two core sources, followed by human editorial review. Sources: TechCrunch (the filing and core arguments), The Verge (litigation background and timeline); original brief language follows the court filing. The text explicitly distinguishes a 'court brief' from a 'ruling' and does not present the government's position as a judicial conclusion.
