特朗普与包括Anthropic的Dario Amodei、OpenAI的Greg Brockman、Nvidia的Jensen Huang、Meta的Mark Zuckerberg、Google的Sundar Pichai和Elon Musk在内的六位科技高管签署了一份人工智能安全协议。2 特朗普将这份协议称为"宪法"，并将人工智能重新命名为"超级智能"。13 该协议旨在推动科技企业自我监管AI开发，包括建立四层控制和审计机制，涵盖内部控制和独立外部审计措施。2
特朗普表示，鉴于美国企业在AI领域"主导中国"，政府无需加强AI护栏。3 他拒绝了推行联邦AI监管的必要性，称政府应该让"世界上最聪明的人相互监督"。3 特朗普还声称，之前关于AI安全风险是民主党制造谎言的说法已不再适用，因为他已将该技术改名。3 特朗普表示"我认为我看到了巨大的自我监管"。2
不过，这份协议仅具非约束力。2 白宫公布的协议文件中出现了拼写错误，将"United States"误写为"Unites States"。2 民调显示，61%的美国选民反对建设数据中心，71%的美国人认为政府对AI监管不足。2
President Trump has signed an artificial intelligence safety accord with six leading technology executives, framing the initiative as a foundational step toward industry self-regulation rather than federal oversight 2. The agreement emerged following a dinner meeting between Trump and the tech leaders 3, and represents his administration's preferred approach to managing AI development without formal government regulation 3.
The accord was signed by Anthropic CEO Dario Amodei, SpaceX's Elon Musk, OpenAI's Greg Brockman, Nvidia's Jensen Huang, Meta's Mark Zuckerberg, and Google's Sundar Pichai 2. The framework establishes four layers of control and audit mechanisms, including internal controls and independent external audits 2. Trump characterized the arrangement as demonstrating substantial industry commitment to self-monitoring 2, asserting that "the world's smartest people" could effectively oversee one another without government intervention 3. He further claimed that American companies "dominate China" in the AI sector, obviating the need for strengthened federal guardrails 3.
Trump has renamed artificial intelligence "Super Intelligence," a terminological shift he presented as superseding previous characterizations of AI safety concerns 3. In earlier framing, Trump had dismissed warnings about AI risks as Democratic-manufactured falsehoods; he now suggests this critique no longer applies following the rebranding 3. The White House document containing the accord, however, contained a spelling error, rendering "United States" as "Unites States" in the presidential designation 2.
Public sentiment on AI governance diverges sharply from the administration's self-regulatory model: 61 percent of American voters oppose building data centers for AI expansion, and 71 percent believe the government's current AI oversight remains inadequate 2. The agreement itself carries no binding legal force, though it may establish groundwork for future formal legislation 2.
评论
还没有评论，欢迎留下第一条。