特朗普新AI規忽視開源模型 —— 且無法定義「安全風險」
特朗普的自願ai測試框架豁免了開源模型，且缺乏對「前沿」或「國家安全風險」的明確定義，給實驗室留下了監管模糊地帶。
the trump administration established a voluntary ai testing framework that mandates a 30-day federal review of closed-source models prior to release to address cybersecurity concerns. the guidelines explicitly exempt open models, prohibiting restrictions on them after release, yet fail to define what constitutes 'state-of-the-art' capabilities or a 'national security risk.' major labs like openai and anthropic attended white house briefings seeking clarity, but the vague definitions leave smaller providers uncertain about compliance standards.