2026年3月18日 美国东部时间上午12:42 / 路透社
作者：杰克·奎因
图片说明：2024年5月20日拍摄的这张插图中可以看到 Anthropic 的标志。路透社/达多·鲁维奇/文件照片 购买授权许可，在新标签页中打开
公司
Anthropic PBC 关注
纽约，3月17日（路透社） – 特朗普政府在周二向法院提交的文件中表示，五角大楼将 Anthropic 列入黑名单的做法是合理且合法的，同时反对这家人工智能实验室对该决定提起的这场高风险诉讼。
3月3日，在 Anthropic 拒绝移除其技术用于自主武器或国内监控的限制措施后，国防部长彼得·赫格斯蒂（Pete Hegseth）将这款广受欢迎的人工智能助手 Claude 的开发商 Anthropic 认定为国家安全供应链风险。
通过《每日 docket》新闻通讯将最新法律新闻直接发送到您的收件箱，开启您的早晨。立即在此注册。
广告 · 滚动继续
特朗普政府在其提交的文件（新标签页打开）中称，Anthropic 声称美国的行动违反美国宪法第一修正案的言论保护规定，但这一主张不太可能成立。该政府方断言，这场纠纷源于合同谈判和国家安全考量，而非报复行为。
政府的法律文件中提到：“只有当 Anthropic 拒绝解除对其产品使用的限制时——而这种拒绝行为本身并非受保护的言论——总统才指示所有联邦机构终止与 Anthropic 的业务关系。” 美国司法部提交的文件称“没有人声称要限制 Anthropic 的表达活动”。
广告 · 滚动继续
Anthropic 在加州联邦法院提起诉讼，要求法官在案件审理期间阻止五角大楼的决定。一些法律专家表示，该公司似乎拥有一个有力的论点，即政府的行为越权了。
Anthropic 在一份声明中表示正在审查政府的提交内容。该公司称“寻求司法审查并不会改变我们长期以来致力于利用人工智能保护国家安全的承诺，但这是保护我们的业务、客户和合作伙伴的必要步骤。”
白宫尚未立即回应置评请求。
唐纳德·特朗普总统支持赫格斯蒂的举措，该举措将 Anthropic 排除在有限的军事合同之外，但据该公司高管称，这可能会损害其声誉，并导致今年损失数十亿美元。
广告即将出现！节点运行失败下一个 保持下一个关闭英语 480p 自动 (480p) 关于 Connatix V243956375 关于 Connatix V243956375 1/1 跳过广告 广告后继续观看！节点运行失败访问广告商网站[前往页面]
这一认定是在五角大楼与 Anthropic 经过数月谈判陷入僵局之后做出的，特朗普和赫格斯蒂借此指责该公司，并指控其因使用限制而危及美国人的生命安全。
Anthropic 对这些指控表示异议，并表示人工智能目前尚未安全到可以用于自主武器。该公司原则上反对国内监控。
在3月9日提起的诉讼中，Anthropic 称这一“前所未有的且非法的”认定侵犯了其言论自由和正当程序权利，同时违反了联邦机构在做决策时必须遵循特定程序的法律规定。
五角大楼还根据另一项法律将 Anthropic 认定为供应链风险，这可能会将该限制扩大到整个政府层面。
Anthropic 正在华盛顿特区上诉法院就这一决定提起第二次诉讼。
杰克·奎因在纽约报道；迈克·斯卡尔塞拉在华盛顿补充报道；林肯·费斯特编辑
我们的标准：路透社信托原则（新标签页打开）
Trump administration defends Anthropic blacklisting in US court
March 18, 2026 12:42 AM UTC / Reuters
By Jack Queen
节点运行失败
Anthropic logo is seen in this illustration taken May 20, 2024. REUTERS/Dado Ruvic/File Photo Purchase Licensing Rights, opens new tab
Companies
Anthropic PBC Follow
NEW YORK, March 17 (Reuters) – The Trump administration said in a Tuesday court filing that the Pentagon’s blacklisting of Anthropic was justified and lawful, opposing the artificial intelligence lab’s high-stakes lawsuit challenging the decision.
Defense Secretary Pete Hegseth designated Anthropic, the maker of popular AI assistant Claude, a national security supply chain risk on March 3 after the company refused to remove guardrails against its technology being used for autonomous weapons or domestic surveillance.
Jumpstart your morning with the latest legal news delivered straight to your inbox from The Daily Docket newsletter. Sign up here.
Advertisement · Scroll to continue
The Trump administration’s filing, opens new tab says Anthropic is unlikely to succeed on its claims that the U.S. action violated speech protections under the U.S. Constitution’s First Amendment, asserting the dispute stems from contract negotiations and national security concerns, not retaliation.
“It was only when Anthropic refused to release the restrictions on the use of its products — which refusal is conduct, not protected speech — that the President directed all federal agencies to terminate their business relationships with Anthropic,” the administration’s legal filing said. The filing, from the U.S. Justice Department, said “no one has purported to restrict Anthropic’s expressive activity.”
Advertisement · Scroll to continue
Anthropic’s lawsuit in California federal court asks a judge to block the Pentagon’s decision while the case plays out. Some legal experts say the company appears to have a strong case that the government overreached.
In a statement, Anthropic said it was reviewing the government’s filing. The company said “seeking judicial review does not change our longstanding commitment to harnessing AI to protect our national security, but this is a necessary step to protect our business, our customers, and our partners.”
The White House did not immediately respond to a request for comment.
President Donald Trump backed Hegseth’s move, which excludes Anthropic from a limited set of military contracts but could damage the company’s reputation and cause billions of dollars in losses this year, according to its executives.
Ad Break Coming Up!节点运行失败NEXT Stay Next Off English 480p Auto (480p)About Connatix V243956375 About Connatix V243956375 1/1 Skip Ad Continue watching after the ad!节点运行失败Visit Advertiser website[GO TO PAGE]
The designation came after months of negotiations between the Pentagon and Anthropic reached an impasse, prompting Trump and Hegseth to denounce the company and accuse it of endangering American lives with its usage restrictions.
Anthropic has disputed those claims and said AI is not yet safe enough to be used in autonomous weapons. The company said it opposes domestic surveillance as a matter of principle.
In its March 9 lawsuit, Anthropic said the “unprecedented and unlawful” designation violated its free speech and due process rights, while running afoul of a law requiring federal agencies to follow specific procedures when making decisions.
The Pentagon separately designated Anthropic a supply chain risk under a different law that could expand the order to the entire government.
Anthropic is challenging that move in a second lawsuit in a Washington, D.C. appeals court.
Reporting by Jack Queen in New York. Additional reporting by Mike Scarcella in Washington; Editing by Lincoln Feast
Our Standards: The Thomson Reuters Trust Principles., opens new tab