美国联邦调查局、国家安全局和网络安全基础设施安全局周二联合发布网络安全建议，指控中国AI开发商自至少2024年末以来，大规模"蒸馏"美国前沿AI模型 1。被指控的企业包括DeepSeek、阿里巴巴、Moonshot AI和Z.ai，美国方面声称这些公司从Anthropic的Claude、OpenAI的GPT、Google的Gemini和SpaceX AI的Grok等模型中进行蒸馏，相关活动"可能得到中国政府知晓" 1。
中国外交部随即回应这一指控。发言人毛宁表示，"中国AI发展是高水平科技自立自强的结果"，呼吁两国作为AI大国应加强合作 1。与此同时，美国财政部长斯科特·贝森特则强调，"中国蒸馏我们的模型，他们永远无法超越我们" 1。
这一争端预计将成为特朗普与习近平本月计划会谈的议题之一 1。
The U.S. Federal Bureau of Investigation, National Security Agency, and Cybersecurity and Infrastructure Security Agency released a joint advisory on Tuesday accusing Chinese AI developers of conducting large-scale distillation of American frontier AI models since at least the end of 2024 1. The targeted models included Anthropic's Claude, OpenAI's GPT, Google's Gemini, and SpaceX AI's Grok 1. The accused Chinese companies were identified as DeepSeek, Alibaba, Moonshot AI, and Z.ai, with the activity characterized as potentially coordinated with Chinese government knowledge 1.
China's Foreign Ministry spokeswoman Mao Ning rejected the accusations, stating that "China's AI development is the result of high-level technological self-reliance and self-strengthening" 1. She called for the two nations to strengthen cooperation as major AI powers 1. U.S. Treasury Secretary Scott Bessent dismissed Beijing's capabilities, asserting "China distills our models, but they will never surpass us" 1.
The dispute is expected to feature in planned talks between Trump and Xi Jinping scheduled for later this month, as AI governance emerges as a point of contention between the two countries 1.
评论
还没有评论，欢迎留下第一条。