---
title: "Pentagon Considers Ending Partnership with AI Firm Anthropic Over Usage Restrictions"
description: "The U.S. Pentagon is considering ending its partnership with AI firm Anthropic due to the company's restrictions on military use of its models. The Pentagon seeks broader access for military applicati"
type: "news"
locale: "en"
url: "https://longbridge.com/en/news/275992561.md"
published_at: "2026-02-15T02:36:06.000Z"
---

# Pentagon Considers Ending Partnership with AI Firm Anthropic Over Usage Restrictions

> The U.S. Pentagon is considering ending its partnership with AI firm Anthropic due to the company's restrictions on military use of its models. The Pentagon seeks broader access for military applications, including weapons development and intelligence gathering, but Anthropic insists on prohibiting mass surveillance of U.S. citizens and fully autonomous weapons. Frustration has grown after months of negotiations, and the Pentagon may reduce collaboration or terminate the partnership if a suitable replacement is found.

The U.S. Pentagon is contemplating ending its partnership with Anthropic, an artificial intelligence company, due to the firm's insistence on imposing certain restrictions on military use of its models. According to Jin10, a senior government official disclosed that the Pentagon is urging four leading AI labs to allow military use of their tools for 'all legitimate purposes,' including sensitive areas such as weapons development, intelligence gathering, and battlefield operations. Anthropic has not agreed to these terms, and after months of difficult negotiations, the Pentagon has grown frustrated. Anthropic maintains that two areas must remain prohibited: mass surveillance of U.S. citizens and fully autonomous weapon systems. The senior official noted that there is considerable ambiguity regarding what should fall into these categories and what should not. Moreover, negotiating each specific use case with Anthropic or dealing with Claude unexpectedly blocking certain applications is not feasible for the Pentagon. The official stated that 'anything is possible,' including temporarily reducing collaboration with Anthropic or completely terminating the partnership. 'But if we believe this is the right course of action, we must find a suitable replacement for them,' the official added.

## Related News & Research

| Title | Description | URL |
|-------|-------------|-----|
| 美國戰爭部下最後通牒 威脅終止 Anthropic 軍方合同 另 Anthropic 放寬核心安全政策 | 美國戰爭部向 Anthropic 發出最後通牒，要求其在週五前遵守五角大樓的 AI 模型使用要求，否則將取消合同。Anthropic 為保持競爭力，決定放寬核心安全政策，允許軍方在合法應用場景中使用其 AI 模型。此舉可能影響與軍方合作的多 | [Link](https://longbridge.com/en/news/276826425.md) |
| 俄烏戰爭將滿 4 週年 基輔又傳遭空襲 2 人受傷 | 俄烏戰爭將滿 4 週年，基輔今晨遭彈道武器襲擊，造成 2 人受傷。基輔市軍事管理局發出空襲警報，呼籲民眾避難。市長克裡契科表示，受傷者為一名婦女和一名孩童，已送醫。自 2022 年 2 月俄羅斯全面入侵烏克蘭以來，基輔頻繁遭受飛彈和無人機攻 | [Link](https://longbridge.com/en/news/276537616.md) |
| Anthropic 控 3 家中企竊取 Claude 模型數據 恐涉國安 | 美國人工智慧公司 Anthropic 指控三家中國 AI 公司（深度求索、月之暗面、MiniMax）非法竊取其聊天機器人 Claude 的技術，稱此為工業規模的智慧財產權竊取。Anthropic 表示，這些公司通過 1600 萬次與 Cla | [Link](https://longbridge.com/en/news/276666727.md) |
| Anthropic 拒絕五角大樓要求，特朗普下令聯邦機構停用 Anthropic 產品 | 特朗普聲稱 Anthropic 試圖強迫美國國防部遵守他們的服務條款，而不是美國的憲法，給予聯邦機構六個月過渡期。此前，國防部要求 Anthropic 允許五角大樓不受限制使用 Claude，但該公司堅持拒絕將其用於大規模監控美國公民或全自 | [Link](https://longbridge.com/en/news/277253154.md) |
| 因拒絕開放 “所有權限”，Anthropic 恐遭美國防部封殺，xAI 火速 “補位” | 由於 Anthropic 與美國國防部在 AI 軍事應用倫理條款上存在根本性分歧，五角大樓向其下達最後通牒，要求在週五下午 5:01 前接受 “所有合法用途” 條款，否則將被列入 “供應鏈風險” 黑名單。Anthropic 首席執行官 Da | [Link](https://longbridge.com/en/news/277128206.md) |

---

> **Disclaimer**: This article is for reference only and does not constitute any investment advice.