AI/LLM 使用正成为对开源维护者的“拒绝服务攻击”
英文标题：🔶 AI/LLM Usage Becoming a "Denial of Service Attack" on Open-Source Maintainers
随着 AI/LLM Agent 生成的大量低质量问题报告涌入，开源项目维护者面临的审核负担正在显著增加。Red Hat 虚拟化工程师 Daniel Berrangé 表示，QEMU 在不到 10 分钟内收到同一用户提交的 125 多份 UBSan 相关报告，且多数未遵循模板、缺少人工分析和修复方案。该用户当天共创建了 132 份问题报告，并已因滥用行为被举报至 GitLab。Berrangé 建议 GitLab 对非项目成员设置问题报告创建频率限制，以降低此类 AI 驱动垃圾报告对维护者造成的影响。
35.7 KB
34.8 KB
31.2 KB
中文
围绕 AI / LLM Agent 给开源软件项目维护者带来日益沉重负担的最新批评是：过多的错误报告实际上正在对项目维护者造成一种“拒绝服务”攻击。
Red Hat 虚拟化工程团队的开源 Linux 虚拟化软件工程师 Daniel Berrangé，以目前影响 QEMU 的错误报告涌入现象提出了这一颇具意味的类比。一名用户在短短几分钟内就向 QEMU 提交了 125 多份独立错误报告。最终，这些报告都与未定义行为检测器（Undefined Behavior Sanitizer，UBSan）的输出及断言有关，并没有真正紧急的错误问题。这些报告既未提供补丁或修复方案，也只是低投入的提交。
这位 Red Hat 工程师在 Mastodon 上解释道：
“QEMU 遭到一名报告者的集中提交：在不到 10 分钟内创建了超过 125 份独立错误报告，全部与 UBSan 报告有关。这些报告无视错误报告模板，没有任何迹象表明有人类对 UBSan 输出进行过分析，也没有提出修复方案。”
Daniel 最终得出结论：
“这实际上是对项目维护者的拒绝服务攻击。”
用户 C4oy1zZ / anwardawa981 当天创建的这 132 份错误报告，已作为滥用行为被举报至 GitLab。在这 132 份错误报告中，许多报告的创建间隔仅有数秒。
Daniel 进一步建议，如果报告者并非项目成员，GitLab 应对单个报告者可创建的问题报告数量实施速率限制。
English / 原文
The latest criticism of the increased burden placed on open-source software project maintainers caused by AI / LLM agents is around too many bug reports effectively causing a "denial of service" attack on project maintainers.
Open-source Linux virtualization software engineer Daniel Berrangé of Red Hat's virtualization engineering team brought up this interesting analogy with the influx of bug reports today affecting QEMU. In a matter of minutes were 125+ separate bug reports sent in by one user for QEMU. They ultimately were about Undefined Behavior Sanitizer (UBSan) outputs and assertions without any really pressing bug report. These bug reports were without any proposed patches/fixes and just low effort submissions.
The Red Hat engineer
on Mastodon:
"QEMU has been hit by a single reporter filing over 125 separate bug reports in under 10 minutes, all for ubsan reports. The reports ignored the bug template and showed no sign of any human analysis of the ubsan output and had no proposed fix."
Ultimately, Daniel concluded, "
this is effectively a denial of service attack on the project maintainers.
"
For this user C4oy1zZ / anwardawa981 and the
created today, was reported to GitLab as abusive behavior. With the 132 bug reports, many of them were created simply seconds apart.
Daniel went on to suggest GitLab enact rate limiting the number of bug reports that can be opened by a single reporter if they are not a project member.