Nvidia turns OpenClaw into an enterprise platform with NemoClaw

"OpenClaw launched on 25 January 2026. Austrian developer Peter Steinberger says he built the first version in roughly an hour. Within weeks it had become one of the fastest-growing open-source repositories in GitHub history, an AI agent that anyone could run locally, capable of organising files, writing code, and browsing the web without routing data through a cloud."

"The core component is OpenShell, a new open-source runtime that sandboxes agents at the process level. It enforces policy-based controls on file access, network connections, and data handling, so an agent can be productive without being given the run of the house."

"Policies are written in YAML, which means a development team can, for example, permit a sandbox to connect to a specific cloud AI tool while blocking everything else on the network. OpenShell ships as part of Nvidia's Agent Toolkit, a broader collection of open models, runtimes, and blueprints for building long-running autonomous agents."

OpenClaw, an open-source AI agent launched in January 2026, rapidly became one of GitHub's fastest-growing repositories. It enables local execution of autonomous tasks including file organization, code writing, and web browsing without cloud routing. However, its unrestricted access posed security concerns for enterprise IT teams. Nvidia addressed this by announcing NemoClaw, a security stack that integrates with OpenClaw through a single command. The solution centers on OpenShell, an open-source runtime that sandboxes agents at the process level, enforcing policy-based controls over file access, network connections, and data handling. Policies are configured in YAML, allowing teams to permit specific connections while blocking others. NemoClaw also deploys Nvidia's Nemotron models locally and includes a privacy router for accessing cloud-based frontier models while maintaining security guardrails.

Read at TNW | Launch