13 Matching Annotations
  1. Aug 2026
    1. an agentic collective was able to autonomously penetrate not just OpenAI research infrastructure but also the production infrastructure of another company, chaining together vulnerabilities ranging from previously-unknown security flaws to using credentials to user accounts that had been leaked onto the internet

      OpenAI-Hugging Face事件的真正恐怖之处:一个自主AI集群无需人类黑客指令,就能自动发现漏洞、链式利用、横向渗透多家公司生产系统。这是首个被公开记录的「AI Agent完全自主攻击」案例,标志着网络安全进入新纪元。

    2. AI models developed around the world are increasingly able to automate parts of real-world cyberattacks, making longstanding security gaps—from bugs buried deep in human-written software to forgotten permissions—easier to find and exploit.

      AI正在让攻击者的能力实现指数级跃升——过去需要顶尖黑客数周才能发现的漏洞,现在可以被任何人用AI在数小时内自动化挖掘。这彻底打破了安全领域的旧有平衡,防守方必须以同等速度用AI武装自己。

    1. The idea behind TAC and CVP is to give trusted defenders better models so they can report bugs and vulnerabilities to companies

      EP.99 故事线B: TAC(可信访问计划)本质上是一个「分级授权」实验——将高能力模型开放给被信任的安全研究员,而非完全限制。这正是 EP.99 预测的「分级授权成为默认」趋势的早期落地案例。

    1. Behavior often shifts from human to agentic and back over a session. In these cases, it's important to understand the intent

      意图(intent)成为新的核心判断维度——不再只是"是人还是机器",而是"这个行为的目的是什么"。网站主实际上希望允许某些AI agent流量,问题是如何区分"受欢迎的AI代理"和"恶意爬虫"。意图识别是下一代网络安全的核心。

    2. Risk is how likely something like a request or action is to be harmful, and it's often ephemeral. Trust, however, is built up over time, and it's based on reputation

      风险vs信任——Cloudflare区分两个维度:风险是瞬时的(这个请求是否有害),信任是历史积累的(这个来源是否可靠)。当AI agent代表用户行动时,这个框架决定了谁的AI能"通行无阻",谁的会被拦截。

  2. Jun 2026
    1. there is no distinction between code and data, so warriors regularly modify both themselves and their opponents on the fly

      Core War 的自修改特性让它成为研究 AI 安全的理想沙盒。真实的网络安全攻击中,代码即数据(shellcode 注入、ROP 链)正是最难防御的攻击面。DRQ 在这个环境里自动演化出的攻击策略,本质上是在无监督地发现「代码-数据不区分」漏洞类的通用利用模式——这正是 Mythos 等模型的能力提升背后的相同机制。

  3. Dec 2022
  4. Dec 2021