Anthropic Resumes Pentagon Talks to Avoid Supply Chain Risk Designation

Anthropic urgently restarts negotiations with the Pentagon to avoid being listed as a U.S. defense supply chain risk enterprise. The core dispute focuses on data usage clauses and AI ethical boundaries. The company has already secured a $200 million defense contract and entered a classified environment.

According to a report by the Financial Times, Anthropic CEO Dario Amodei recently urgently restarted negotiations with the U.S. Department of Defense in an attempt to prevent the company from being formally listed as a "supply chain risk" enterprise during the Trump administration, which would deprive it of the opportunity to participate in defense projects. The core of these negotiations revolves around a key clause – regarding "bulk data acquisition analysis rights." Anthropic insists on retaining this clause, stating that it is intended to prevent the AI models it develops from being used for large-scale domestic surveillance. The company has also explicitly prohibited its technology from being used in lethal autonomous weapon systems, demonstrating a high degree of vigilance regarding ethical boundaries.

Anthropic Resumes Pentagon Talks to Avoid Supply Chain Risk Designation插图
Although the Pentagon had suggested that it could accept most of the other terms of cooperation if Anthropic removed the data usage clause, the company ultimately refused to concede. External analysis suggests that this position reflects Anthropic's trade-off between technological ethics and commercial interests. It is worth noting that Anthropic has established deep ties with the U.S. defense system. In July 2025, the company was awarded a contract worth up to $200 million by the Department of Defense, becoming the first AI vendor to receive a national-level classified environment usage license, and its models have been used for sensitive national security missions.
Anthropic Resumes Pentagon Talks to Avoid Supply Chain Risk Designation插图1
At the same time, Amodei publicly criticized competitor OpenAI and its CEO Sam Altman, calling their cooperation with the Department of Defense merely a "safety show," implying that the latter lacks substantive constraints on safety and compliance. Currently, both parties are still seeking a compromise. If an agreement is not reached, not only will Anthropic face the risk of contract restrictions, but its partners may also be forced to interrupt collaboration, which will affect the layout of the U.S. defense AI ecosystem.

0 comment A文章作者 M管理员
    No Comments Yet. Be the first to share what you think
Profile
Search
🇨🇳Chinese🇺🇸English