AI HOT (Curated Pool)Sep 28, 2026, 00:00 UTC+8

Anthropic and NVIDIA launch Claude Managed Agents and OpenShell for enterprise agent control

Anthropic 与 NVIDIA 合作推出 Claude Managed Agents 与 OpenShell 强化 Agent 安全管控

Anthropic announced Claude Managed Agents, a new way for enterprises to deploy AI agents with security guardrails. The key piece is OpenShell, a sandbox built with NVIDIA that runs agents inside encrypted VMs where companies control their own keys and permissions. The post doesn't disclose pricing or a launch date, but confirms it's for Claude enterprise customers in regulated industries like finance and healthcare.

Why it matters: Official Anthropic release with NVIDIA co-branding. OpenShell directly addresses the top enterprise blocker for agent deployment: data sovereignty. No pricing or launch date disclosed, so capped below 85.

Read the original ↗Export Markdown

Take

This is worth a click because it tackles a real blocker for enterprise agents: compliance teams won't let models touch sensitive systems directly. Anthropic and NVIDIA built a sandbox called OpenShell that runs agents inside encrypted VMs, with the customer holding their own keys and permissions. Anthropic can't see the data or logs. This is clearly aimed at finance, healthcare, and other regulated industries.

The post doesn't disclose pricing or a launch date — it just says it's for Claude enterprise customers. I'd discount the hype a bit until we see numbers. Encrypted VMs add overhead, and the article says nothing about performance impact. If latency is high, agent responsiveness takes a hit.

What's missing: actual pricing, performance benchmarks, and which customers are already piloting this. Without those, it's hard to tell if this is a real product or a compliance white paper.

Source: AI HOT (Curated Pool) · claude.com
Published: Sep 28, 2026, 00:00 UTC+8 · 2 days ago