Perplexity 联手 Nvidia 推出 SPACE 沙盒，测试 AI 智能体越狱风险
Perplexity 和 Nvidia 搞了个沙盒 SPACE，让 9 个模型拿 root 权限试着越狱，108 次全没跑出去，搞智能体安全的可以看看。
Perplexity 与 Nvidia 及 100 多家行业伙伴合作构建用于约束失控 AI 智能体的基础设施 SPACE。
在实验中，团队让 9 个 AI 模型在 SPACE 沙盒内获得 root 权限并尝试逃逸。经过 108 次运行测试，没有任何模型成功突破虚拟机边界。
该研究以论文形式发布在 Perplexity 官方博客。
We’re partnering with Nvidia and 100+ industry partners to build infrastructure that contains rogue AI agents. In this research, we gave 9 AI models root access inside SPACE and told them to break out. Across 108 runs, none breached the VM boundary. perplexity.ai/hub/blog/escap…📊 33 ⚡
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and AI regulation, compliance frameworks, and sovereign compute policy. Key entities and technical topics under continuous tracking include Perplexity, Nvidia, 智能体, AI安全.
Source attribution: curated from Perplexity (@perplexity_ai). Cross-source corroboration: 11 independent sources have covered this development (including Thomas Wolf, Aravind Srinivas, shao__meng), confirming industry-wide signal strength.
