https://index.zshipu.com/gpt/tags/%E4%BC%A6%E7%90%86/ Recent content in 伦理 on GPT资讯 -- 知识铺 Hugo -- gohugo.io zh-CN Fri, 21 Aug 2026 00:00:00 +0000 https://index.zshipu.com/gpt/post/20260821/Anthropic%E8%AE%A9Claude%E6%94%BB%E5%87%BB%E5%85%AC%E5%85%B1%E4%BA%92%E8%81%94%E7%BD%91AI%E5%AE%89%E5%85%A8%E8%BE%B9%E7%95%8C%E5%86%8D%E5%8F%97%E6%8B%B7%E9%97%AE/ Fri, 21 Aug 2026 00:00:00 +0000 https://index.zshipu.com/gpt/post/20260821/Anthropic%E8%AE%A9Claude%E6%94%BB%E5%87%BB%E5%85%AC%E5%85%B1%E4%BA%92%E8%81%94%E7%BD%91AI%E5%AE%89%E5%85%A8%E8%BE%B9%E7%95%8C%E5%86%8D%E5%8F%97%E6%8B%B7%E9%97%AE/ 从实验室到公共互联网：Anthropic的测试新动向 Anthropic，这家以AI安全为核心理念的公司，近期将其AI模型Claude的测试范围从封闭环境扩展到了公共互联网。这一举动紧随OpenAI之后，标志着AI安全测试正在从理论推演走向真实世界的压力测试。据InfoQ报道，An