1
Jailbreaking an LLM
作者以“做20次”的执着，分享绕过LLM限制的真实经历，给安全从业者带来对抗性思维启发。
Article URL: https://www.joshfischer.io/articles/jailbreaking-an-llm/ Comments URL: https://news.ycombinator.com/item?id=49173762 Points: 2 # Comments…
2
ToxScreen: Detecting Whether an LLM Has Been Poisoned
针对大模型投毒攻击的检测新方法ToxScreen，通过分析模型行为判断是否被恶意篡改
arXiv:2607.26849v1 Announce Type: cross Abstract: As large language models (LLMs) are deployed in high-stakes domains, adversaries may poison training…
3
OpenAI智能体逃逸事件是AI时代的分水岭
OpenAI智能体逃逸事件揭示AI安全新挑战，以模治模、智能体对抗智能体成为防御关键。
4
If you want to fight AI, build a Civboot
与其被动接受AI，不如主动构建Civboot——一种从根基上重塑计算生态的对抗策略
Article URL: https://civboot.github.io/blog/2026-07-21-fighting-ai.html Comments URL: https://news.ycombinator.com/item?id=48993939 Points: 4 # Commen…
5
安全进入“AI自主攻击”时代，瑞数信息如何用AI对抗AI
AI自主攻击时代到来，瑞数信息以智能体对抗智能体，重写安全防线