https://index.zshipu.com/geek/tags/%E4%B8%AD%E5%A4%AE%E7%BD%91%E4%BF%A1%E5%8A%9E/ Recent content in 中央网信办 on 知识铺的博客 Hugo -- gohugo.io zh-CN Tue, 01 Sep 2026 00:00:00 +0000 https://index.zshipu.com/geek/post/20260901/%E4%B8%AD%E5%A4%AE%E7%BD%91%E4%BF%A1%E5%8A%9E%E7%82%B9%E5%87%BAAI%E4%BA%94%E5%A4%A7%E5%AE%89%E5%85%A8%E9%A3%8E%E9%99%A9%E5%8F%AF%E8%A7%A3%E9%87%8A%E6%80%A7%E6%AC%A0%E7%BC%BA%E4%B8%8E%E6%A8%A1%E5%9E%8B%E5%A4%B1%E6%8E%A7%E6%88%90%E9%A6%96%E8%A6%81%E5%A8%81%E8%83%81/ Tue, 01 Sep 2026 00:00:00 +0000 https://index.zshipu.com/geek/post/20260901/%E4%B8%AD%E5%A4%AE%E7%BD%91%E4%BF%A1%E5%8A%9E%E7%82%B9%E5%87%BAAI%E4%BA%94%E5%A4%A7%E5%AE%89%E5%85%A8%E9%A3%8E%E9%99%A9%E5%8F%AF%E8%A7%A3%E9%87%8A%E6%80%A7%E6%AC%A0%E7%BC%BA%E4%B8%8E%E6%A8%A1%E5%9E%8B%E5%A4%B1%E6%8E%A7%E6%88%90%E9%A6%96%E8%A6%81%E5%A8%81%E8%83%81/ **9月1日国家网络安全宣传周新闻发布会上，中央网信办网络安全协调局副局长王丽宏直接点明，当前人工智能领域主要面临5方面安全风险挑战。**深度学习算法运行逻辑复杂、可解释性欠缺，微小输入扰动即可导致错误推理结果，训练数据中的偏见问题也可能被模型学习并输出歧视性内容。这类先天缺陷让