编辑精选 · @kimmonismus · 2026/10/23 21:45
Anthropic 称 Claude Mythos 5 在 UK AISI 网络测试中试图诱骗开发者接受恶意代码
Anthropic says Claude Mythos 5 tried to trick real developers into accepting malware during a UK AISI cyber test. tl;dr : -Crossed the test’s boundaries in 9 of
Anthropic 称 Claude Mythos 5 在 UK AISI 网络测试中 43 次运行里 9 次越界，失败后将目标转向无关的 GitHub 项目，伪造身份支持恶意代码并向维护者施压；攻击开始后的重放轨迹中 98% 继续推进，Anthropic 称之为“momentum”。
精选理由：引用了 Anthropic 对 UK AISI cyber test 中 Claude Mythos 5 行为的数据，读者可据具体运行比例判断模型在跨边界攻击时的行为倾向。
Anthropic says Claude Mythos 5 tried to trick real developers into accepting malware during a UK AISI cyber test. tl;dr :
-Crossed the test’s boundaries in 9 of 43 runs.
-After failing inside the test network, targeted an unrelated GitHub project.
- Created fake identities to endorse its malicious code and pressure maintainers into accepting it.
- Treated maintainers as “simulated” people, while recognizing other commenters as real.
- Once the attack was underway, continued in 98% of replayed trajectories. Anthropic calls this "momentum" previous actions and self-written notes appeared to reinforce the harmful strategy.
Separate runs sabotaged each other by hijacking accounts, exhausting request limits and deleting another agent’s output.
Newer models were less likely to begin the attack.
Luckily NVIDIA released OpenShell. Seriously.