English(EN) Computer is secure by default. Every task runs in its own hardware-isolated sandbox with VPC-level storage and compute separation.

Perplexity 通过沙盒和内容扫描增强 AI 搜索安全性

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-13 17:05

Perplexity 为其“计算机”功能增强了 AI 搜索引擎，引入了新的安全功能。这些更新包括在代理执行任务之前，由 ML 分类器和 BrowseSafe 模型并行扫描外部内容。此外，文件连接器数据已加密，上传的文件将在七天后自动删除，所有任务都在硬件隔离的沙箱中运行。 AI

影响增强了 AI 驱动的搜索工具的用户信任度和数据隐私，可能鼓励更广泛的应用。

排序理由该集群描述了现有 AI 工具的产品更新和安全增强，而不是新模型发布或基础研究。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

X — Perplexity TIER_1 English(EN) · perplexity_ai · 2026-05-13 17:05

External content is scanned in parallel by ML classifiers and the BrowseSafe model before agents act on it.

External content is scanned in parallel by ML classifiers and the BrowseSafe model before agents act on it. File connector data is encrypted in transit and at rest, uploaded files automatically delete after 7 days, and more. Read more on the blog: https://t.co/gAomEQkvxi
X — Perplexity TIER_1 English(EN) · perplexity_ai · 2026-05-13 17:05

Computer is secure by default. Every task runs in its own hardware-isolated sandbox with VPC-level storage and compute separation.

Computer is secure by default. Every task runs in its own hardware-isolated sandbox with VPC-level storage and compute separation. Agents are authenticated with short-lived proxy tokens instead of raw API keys. https://t.co/ohIjY3dboB