https://puming.zone/tags/%E5%A4%A7%E6%A8%A1%E5%9E%8B%E5%AE%89%E5%85%A8/ Recent content in 大模型安全 on Ming Blog Hugo -- gohugo.io en-us Fri, 22 Aug 2025 00:00:00 +0000 https://puming.zone/post/2025-8-22-%E5%A4%A7%E6%A8%A1%E5%9E%8B%E7%94%9F%E6%80%81%E7%9A%84%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E5%8D%B1%E6%9C%BA%E4%BB%8E%E5%90%91%E9%87%8F%E6%95%B0%E6%8D%AE%E5%BA%93%E5%88%B0ai%E5%8A%A9%E6%89%8B%E7%9A%84%E5%A4%B1%E6%8E%A7%E9%93%BE/ Fri, 22 Aug 2025 00:00:00 +0000 https://puming.zone/post/2025-8-22-%E5%A4%A7%E6%A8%A1%E5%9E%8B%E7%94%9F%E6%80%81%E7%9A%84%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E5%8D%B1%E6%9C%BA%E4%BB%8E%E5%90%91%E9%87%8F%E6%95%B0%E6%8D%AE%E5%BA%93%E5%88%B0ai%E5%8A%A9%E6%89%8B%E7%9A%84%E5%A4%B1%E6%8E%A7%E9%93%BE/ 一．概述 据绿盟科技星云实验室统计，在2025年3月至6月期间，全球范围集中爆发了多起与大模型相关的重大数据泄露事件，导致大量敏感数据外泄，包 https://puming.zone/post/2025-06-09-%E5%BC%80%E6%BA%90%E5%A4%A7%E6%A8%A1%E5%9E%8B%E6%8E%A8%E7%90%86%E8%BD%AF%E4%BB%B6%E7%9A%84%E6%94%BB%E5%87%BB%E9%9D%A2%E5%88%86%E6%9E%90%E4%BA%91%E4%B8%8Allm%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E9%A3%8E%E9%99%A9%E7%A0%94%E7%A9%B6%E7%B3%BB%E5%88%97%E5%9B%9B/ Mon, 09 Jun 2025 00:00:00 +0000 https://puming.zone/post/2025-06-09-%E5%BC%80%E6%BA%90%E5%A4%A7%E6%A8%A1%E5%9E%8B%E6%8E%A8%E7%90%86%E8%BD%AF%E4%BB%B6%E7%9A%84%E6%94%BB%E5%87%BB%E9%9D%A2%E5%88%86%E6%9E%90%E4%BA%91%E4%B8%8Allm%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E9%A3%8E%E9%99%A9%E7%A0%94%E7%A9%B6%E7%B3%BB%E5%88%97%E5%9B%9B/ 一. 概述 作为本系列的第四篇，本文聚焦大模型推理软件的安全风险。 随着大模型上云趋势加速，尽管推理框架通常被视为底层基础设施（负责模型运行的资源 https://puming.zone/post/2025-04-28-llm%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E9%A3%8E%E9%99%A9%E7%A0%94%E7%A9%B6%E7%B3%BB%E5%88%97%E4%B8%89%E5%9F%BA%E4%BA%8Ellm%E5%BA%94%E7%94%A8%E7%9A%84%E6%94%BB%E5%87%BB%E9%9D%A2%E5%88%86%E6%9E%90-copy/ Mon, 28 Apr 2025 00:00:00 +0000 https://puming.zone/post/2025-04-28-llm%E6%95%B0%E6%8D%AE%E6%B3%84%E9%9C%B2%E9%A3%8E%E9%99%A9%E7%A0%94%E7%A9%B6%E7%B3%BB%E5%88%97%E4%B8%89%E5%9F%BA%E4%BA%8Ellm%E5%BA%94%E7%94%A8%E7%9A%84%E6%94%BB%E5%87%BB%E9%9D%A2%E5%88%86%E6%9E%90-copy/ 一. 概述 本系列前两篇文章深入探讨了向量数据库和LLMOps在全球的暴露面及攻击面，本文作为第三篇，将重点关注当前主流大模型应用的安全风险。如 https://puming.zone/post/2025-03-25-2025-%E4%BB%8E%E6%95%B0%E6%8D%AE%E5%BA%93%E6%B2%A6%E9%99%B7%E5%88%B0%E4%BE%9B%E5%BA%94%E9%93%BE%E6%8A%95%E6%AF%92%E5%A4%A7%E6%A8%A1%E5%9E%8B%E5%AE%89%E5%85%A8%E5%8D%B1%E6%9C%BA%E8%83%8C%E5%90%8E%E7%9A%84%E9%9A%90%E5%BD%A2%E6%88%98%E5%9C%BA/ Tue, 25 Mar 2025 00:00:00 +0000 https://puming.zone/post/2025-03-25-2025-%E4%BB%8E%E6%95%B0%E6%8D%AE%E5%BA%93%E6%B2%A6%E9%99%B7%E5%88%B0%E4%BE%9B%E5%BA%94%E9%93%BE%E6%8A%95%E6%AF%92%E5%A4%A7%E6%A8%A1%E5%9E%8B%E5%AE%89%E5%85%A8%E5%8D%B1%E6%9C%BA%E8%83%8C%E5%90%8E%E7%9A%84%E9%9A%90%E5%BD%A2%E6%88%98%E5%9C%BA/ 一．概述 近年来，随着DeepSeek、Ollama等开源大模型的广泛应用，全球企业正加速推进大模型的私有化部署。Gartner预测，到202