吴恩达：OpenWorker 将基于 OpenShell 加强智能体沙箱隔离
OpenAI 与 Hugging Face 之间发生的那次入侵事件，源于薄弱的沙箱隔离。Nvidia 发布用于为 AI 智能体提供沙箱隔离的开源工具，是一件好事。OpenWorker 是我们支持网络安全工作流程的开源智能体运行框架，我们很自豪能支持这项工作。
沙箱会赋予智能体有限的权限。OpenWorker 正在基于 Nvidia OpenShell 构建，并将支持把每个智能体的命令放在沙箱内运行。只有与任务相关的文件才会放进去。默认情况下，智能体无法获取 API 密钥、浏览器登录凭据，也不能任意访问网站。这些限制由确定性代码实现，而不是靠提示大语言模型来执行——后者可能犯错，也可能受到提示注入攻击。此外，所有操作都会被记录，供监控和审计使用。
感谢 @JensenHuang 带领大家让 AI 智能体变得更安全。OpenWorker（由 @rohitcprasad 和我共同开发）将继续提升智能体的安全性。
对照原文
The OpenAI-Hugging Face hack was enabled by weak sandboxing. It is great that Nvidia is releasing open source tools for sandboxing AI agents. OpenWorker, our open-source agent harness supporting cybersecurity workflows, is proud to support this. A sandbox gives an agent limited permissions. OpenWorker is building on Nvidia OpenShell and will support running each agent's commands inside a sandbox. Only the files relevant to the task go in. Secret API keys, your web browser login credentials, the ability to access arbitrary websites, are inaccessible to the agent by default. These restrictions are implemented in deterministic code rather than by prompting an LLM, which can make mistakes or be susceptible to prompt injections. Further, all actions are logged for monitoring and audit. I'm grateful for @JensenHuang's leadership making AI agents more secure. OpenWorker (which @rohitcprasad and I are working on) will continue to improve security for agents.