Netlify Moves Every Build to Dedicated Firecracker MicroVMs for Hardware-Level Isolation.

Netlify announced that every build on its platform now runs inside a dedicated Firecracker MicroVM. Each MicroVM gets its own kernel and memory space, is created for a single build, and is destroyed immediately when the build completes.

This is hardware-level isolation, not container-level. And it applies to every build by default, not as an enterprise add-on.

The change addresses a fundamental challenge in running arbitrary code from thousands of different projects at scale: preventing state leaks, eliminating shared containers across projects, and blocking interference between different customers' code.

Firecracker is an open-source virtualization technology originally developed by AWS for Lambda and Fargate. Unlike traditional containers that share a host kernel, each Firecracker MicroVM runs its own kernel instance.

This means builds are isolated at the hardware level. If one build attempts to access another's memory or state, the hypervisor blocks it at the kernel boundary, not just at the process level.

Netlify's decision to make this the default architecture, rather than tiering it as a premium security feature, reflects the company's position on running untrusted code. When you're processing builds from countless repositories, each potentially pulling dependencies and running scripts from third parties, the attack surface is significant.

Nothing, operationally. Builds continue to work exactly as before. The MicroVM layer is invisible to the build process itself.

The isolation model means:

• No environment variables, secrets, or build artifacts can leak between projects
• Dependencies installed in one build cannot affect another
• Malicious code in one build cannot probe or interfere with another customer's infrastructure

Each VM is ephemeral. It exists only for the duration of a single build, then disappears. This stateless approach eliminates entire classes of persistence-based attacks.

The announcement comes as Netlify continues to expand its platform for AI-assisted development. Earlier this month, the company positioned Agent Experience as a bridge from AI setup to production, emphasizing infrastructure that can handle autonomous deployment workflows.

Running code generated by AI agents or pulled from unfamiliar repositories requires strong isolation guarantees. MicroVMs provide that boundary without requiring developers to change how they write or deploy code.

The move also aligns with Netlify's broader security stance. Last month, the company clarified that it does not use customer code for AI training, addressing growing concerns about how platforms handle proprietary repositories and build artifacts.

Other platforms in the deployment space, including Vercel and Railway, use various isolation strategies for build and runtime environments. Netlify's approach puts hardware-level separation at the foundation rather than treating it as an optional layer.

For developers, the security improvement is automatic. For platform operators, it represents a significant architectural commitment to isolation as the default posture.