nFOX Lab launches platform, to open-source Blockhouse sandbox

nFOX Lab emerged from stealth this week with a pitch aimed at companies that want AI capabilities but hesitate to send sensitive information to external providers. Running models internally gives enterprises control over their data, but it also makes them responsible for securing the software and infrastructure that support those models. nFOX's answer is Blockhouse, a sandbox that operates beneath the application layer and uses Linux security mechanisms to remove capabilities rather than instruct models on what to avoid. "You're basically removing the ability to do it in the first place," Allott said.

The system routes communications through Porthole, a monitored channel where requests can be reviewed before they are fulfilled, and keeps runtime credentials and audit controls outside the model process. Blockhouse also checks model integrity: approved files are sealed with SHA-256 hashes before serving, and built-in checks halt serving if those files change. Allott acknowledged the approach does not eliminate human error, as an administrator could mistakenly authorize a malicious destination.

nFOX is self-funded but intends to raise venture capital, and Allott said it has paying beta pilot customers without being profitable yet. He pointed to Nvidia's OpenShell as similar technology focused on securing agents, which nFOX intends to incorporate as its roadmap expands into agent protection over time.