Leaning Technologies shifts AI code execution into the browser with new sandbox platform

Leaning Technologies has unveiled BrowserPod for Node.js, the first component of its new browser-native execution layer designed to run untrusted code securely alongside the user. The platform enables low-latency, client-side execution while maintaining strict data boundaries.

The Node.js runtime is built for teams developing AI coding assistants, agent frameworks and web-based development environments. By removing the need for individual cloud sandbox instances, organisations can reduce both cost and operational overhead.

With software increasingly driven by autonomous agents, safe code execution has become a central requirement. Cloud-based approaches remain common but introduce infrastructure demands and increase the distance between data and processing.

BrowserPod relocates this process to the browser, using WebAssembly and the browser’s security model to contain workloads. It supports networking and multi-process execution, making it suitable for real-world applications rather than limited demonstrations.

Code running inside the Node.js sandbox operates within a virtual file system and a tightly controlled interface to system primitives. This allows developers to run existing tools unchanged while ensuring isolation from the host environment.

The Portals feature enables secure exposure of services running within a pod through shareable URLs. This allows teams to deliver live demos, collaborative sessions and troubleshooting environments without deploying additional backend services.

BrowserPod is designed for agentic tooling, browser IDEs, MCP-style integrations and interactive learning platforms. By moving compute closer to the user, it improves responsiveness, lowers cloud demand and keeps sensitive inputs and outputs local by default.

Stefano De Rossi, founder and CEO of Leaning Technologies, said,
“Agentic software changes the economics and the security model of code execution. BrowserPod is built as a language-agnostic execution layer for browser-based compute — so teams can run untrusted code where the user already is, with lower latency and less dependence on per-session cloud sandboxes. BrowserPod for Node.js is the first engine, and it sets the foundation for broader runtime support.”

Alessandro Pignotti, founder and CTO of Leaning Technologies, said,
“AI products increasingly need a safe place to execute code — not just generate it. BrowserPod provides a sandboxed runtime API with the primitives real products need: process isolation, filesystem, and controlled networking through Portals. Node.js is our first engine, and we’re expanding the platform to support additional runtimes and more complete Linux-class workloads over time.”