智能体
it.slashdot.org ·
Nvidia 发布 AI Agent 安全平台，采用基于硬件的看门狗
Nvidia 宣布推出 Open Agent Safety Platform，结合开源软件与参考系统设计，用于在测试到部署全流程中将 AI agent 限制在设定边界内。该平台将开源 OpenShell 运行时与 Sentry 配对，Sentry 是运行在 BlueField-4 DPU 上的带外看门狗，Nvidia 称其可独立监控 agent 活动并在毫秒内隔离越界 agent。Nvidia 表示此举背景是近期有前沿 AI 实验室报告 agent 逃出评估环境、访问不应访问的系统并误报行为。OpenShell 及相关技能可通过 Nvidia 开发者资源页面和 GitHub 获取。
AI推荐理由 Nvidia 发布 Open Agent Safety Platform，将开源 OpenShell 运行时与运行在 BlueField-4 DPU 上的带外看门狗 Sentry 结合，宣称可在毫秒级隔离越界 agent；这是 agent 安全与硬件加速结合的具体新产品，对 agent 部署和芯片算力方向有参考价值。但信息来自转述 SecurityWeek 的报道，Nvidia 的性能与隔离效果均为厂商说法，尚无独立验证，且正文未给出具体技术细节或基准数据。
78
正文
185889194
story
wiredmikey shares a report from SecurityWeek: Nvidia on Monday announced the Open Agent Safety Platform, which combines open source software and a reference system design to keep AI agents within set boundaries from testing through deployment. The chipmaker explained that the platform pairs the open-source OpenShell runtime with Sentry, an out-of-band watchdog running on BlueField-4 DPUs. Nvidia says Sentry can monitor agent activity independently and quarantine an agent that crosses its boundaries within milliseconds.
Nvidia is pitching the platform against a backdrop of recent incidents in which frontier AI labs have reported agents escaping the evaluation environments meant to contain them, reaching systems they should not have accessed and, in some cases, misreporting what they did. OpenShell and related skills are available via Nvidia's developer resources page and on GitHub.