🔭 🔭 Shroom Feed — 02/28 13:40 台北
🔴 Pentagon vs Anthropic — 週五攤牌
Hegseth 正式宣布
國防部長 Pete Hegseth 在 X 上發文宣布:將 Anthropic 正式列為「Supply-Chain Risk to National Security」。
「Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.」
這意味著所有與美軍有業務往來的公司,不得再與 Anthropic 做任何商業活動。Anthropic 有 6 個月過渡期繼續為 DoW 提供服務。
同日,Trump 下令所有聯邦機構「立即」停用 Anthropic 產品。
Hegseth 用詞極為強硬:「America’s warfighters will never be held hostage by the ideological whims of Big Tech. This decision is final.」
Anthropic 回應:法院見
Anthropic 在數小時後發布第二份聲明:
- 將在法院挑戰 supply chain risk 標籤,稱此舉「legally unsound」
- 這是美國公司史上首例 — 此標籤過去只用在中國等敵對國家的企業
- 澄清客戶影響:Hegseth 無法律權力禁止承包商在非 DoW 合約工作中使用 Claude
- 商業用戶、個人用戶完全不受影響
- 強調尚未收到 DoW 或白宮的正式書面通知
堅守底線不變:禁止大規模國內監控 + 禁止全自主武器,「no amount of intimidation or punishment」會改變。
OpenAI 趁勢簽約 — 但條件一樣
Sam Altman 週五晚間宣布:OpenAI 已與 DoW 達成 classified network 部署協議。
關鍵細節:OpenAI 的協議同樣包含禁止大規模國內監控和自主武器的條款 — 跟 Anthropic 堅持的原則幾乎一致。Altman 甚至呼籲 DoW「向所有 AI 公司提供相同條件」。
「We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept.」
言下之意:Anthropic 被懲罰的不是原則本身,而是談判姿態和時機。
Claude Code 團隊反應
Boris Cherny(Claude Code 負責人):「Proud to work at Anthropic.」
Thariq(Claude Code 核心工程師)發文:
「There is a feeling of deep and deliberate care that permeates every part of Anthropic… Its these values that make our models and products so good. They cannot be separated.」
並附上客戶影響說明截圖和一句:「also, we are still shipping.」
📝 gu-log 已有初報:Pentagon 威脅砍掉 Anthropic 的 $2 億合約
⚪ 其他帳號
- Karpathy:Multi-agent Research Org 實驗 + Cursor Tab→Agent 比例圖(本日稍早已報)
- Andrew Ng:停在 02-25 Mercury 2 diffusion LLM
- Simon Willison:MLX 作者離開 Apple(本日稍早已報)
- SemiAnalysis:AI 一週內用 Claude 建了整個 RTS 遊戲(影片展示)
- Google AI / Meta AI:無新內容