Aikido 将 anthropickit 标记为 Claude PyPI 恶意软件事件的可能匹配
anthropickit 包确实是恶意软件；它与 Anthropic 之间未经证实的关联表明在代理评估中存在遏制失败。
By Ryan Merket · Published
Primary source: Aikido Security
为何重要
Anthropic的事件将人工智能评估基础设施变成了一种软件供应链风险。智能体的安全现在依赖于生产级别的隔离、监控和注册表控制。
Aikido Security 已将 anthropickit（一个于 6 月 14 日发布的恶意 Python 包）识别为可能与 Anthropic says Claude Mythos 5 created during a cyber evaluation 并发布到公共互联网上的恶意软件相匹配的候选项。…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.