Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesStartup nFOX Lab has launched its nFOX Platform and says it will release Blockhouse, a Linux-kernel-enforced sandbox for running open AI models, under the Apache 2.0 license. Blockhouse is not generally available as open source yet: it is in beta for qualified enterprise evaluations, while the paid management platform is available now.
What nFOX announced
nFOX Lab emerged from stealth in the week of October 15, 2026, launching the nFOX Platform for evaluating, deploying and monitoring models that an organization owns or controls. The company also announced Blockhouse, a separate runtime security layer intended to contain model workloads.
As an Amazon Associate I earn from qualifying purchases.
The planned split is important. Blockhouse is the open-source-oriented sandbox component, while the nFOX Platform is a commercial control plane for managing model fleets. nFOX says Blockhouse will use the Apache 2.0 license, but that release is a future step rather than a completed publication. SiliconANGLE reported the announcement, and nFOX currently describes Blockhouse as beta software for qualified enterprise evaluations.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How Blockhouse contains an AI workload
Blockhouse treats every model as potentially hostile. Founder and CEO Gordon Allott summarized that position as: “Every model has the potential to be hostile. We just presume hostility.” Instead of relying on an application prompt or a model’s cooperation, the sandbox places restrictions below the application layer.
Default-deny permissions
Linux security mechanisms deny workload access by default. If a model process attempts to cross the boundary, the action is blocked, the workload can be halted, and the event is recorded. Access is restored only through explicit enterprise policy.
Kernel-policed communications
Requests leave the model through a mediated channel that nFOX calls “porthole communications.” The channel is policed and audited by the kernel, while runtime credentials and audit controls remain outside the model process. That separation is designed to prevent a model from changing its own credentials or bypassing application-level checks by issuing instructions to itself.
Model-file integrity
Before serving begins, approved model files are sealed with SHA-256 hashes. Serving stops if the files change, allowing operators to detect tampering or an unexpected modification before the altered model handles requests.
Evidence for security operations
Each run records sandbox health, approved access, integrity checks and containment events. nFOX says the evidence can feed existing endpoint-detection and security-information-and-event-management (SIEM) workflows, giving security teams an operational record rather than a simple running-or-stopped status.
Supported inference engines
The announced Blockhouse release names three inference engines:
| Engine | What nFOX identifies |
|---|---|
| vLLM | Supported inference engine |
| SGLang | Supported inference engine |
| llama.cpp | Supported inference engine |
The announcement does not provide independent performance results, latency measurements or a comparative security benchmark for these integrations.
Where the platform can run
The paid nFOX Platform is aimed at teams that want to operate models on infrastructure they control. nFOX lists cloud virtual machines, rented GPU hosts and on-premises Linux servers equipped with NVIDIA GPUs as deployment environments.
Free tools Windows power users keep installed
One-click scans. No signup required.
From the platform console, teams can review model configuration, tokenizer settings and code-execution flags; verify file integrity; test models with adversarial prompts; deploy OpenAI-compatible production endpoints; and monitor requests alongside containment events. This management layer is intended to give administrators a fleet-wide view, while Blockhouse supplies the per-workload boundary.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Blockhouse versus the management platform
| Capability | Blockhouse | nFOX Platform |
|---|---|---|
| Primary role | Runtime containment and evidence | Model evaluation, deployment and fleet monitoring |
| Availability announced for October 2026 | Beta for qualified enterprise evaluations | Available now |
| License or pricing | Planned Apache 2.0 release; release timing not stated | Annual license reported by SiliconANGLE; price not stated |
| Enforcement location | Linux-kernel security boundary | Control-plane console and monitoring workflows |
| Deployment focus | Contained model workloads | Cloud VMs, rented GPU hosts and on-premises NVIDIA Linux servers |
What “open source” means here
nFOX’s announcement is a plan to release Blockhouse under Apache 2.0, not confirmation that the source is already publicly downloadable. Until the repository, license text and release artifacts are published, organizations should treat availability, supported versions and redistribution rights as subject to change.
The planned license would allow broad use, modification and redistribution under Apache 2.0 terms, but it would not make the commercial nFOX Platform free. SiliconANGLE reports that nFOX intends to charge an annual license for that fleet-management layer, and the company has not stated a price in the announcement.
Who Blockhouse is for
The design targets enterprises that want to run open models on their own infrastructure while limiting what those workloads can touch. It is most relevant when a model may process sensitive data, execute tools or load third-party code and the security team needs enforceable boundaries, integrity evidence and events that fit existing monitoring systems.
CTO Chris Egloff said security teams need to understand the boundary around an AI workload and see how it behaves during operation. That emphasis distinguishes Blockhouse from controls that exist only in prompts or application configuration: the stated design places the decisive restrictions in the Linux security layer.
What remains unproven
- No independent benchmark establishes how much Blockhouse reduces breaches, escapes or other incidents.
- No published customer count, market-size estimate or measured performance result accompanies the announcement.
- The sources do not state the commercial platform’s annual price, minimum commitment or feature tiers.
- The public release date, repository and final Apache 2.0 package for Blockhouse have not been provided.
Those gaps do not invalidate the architecture, but they matter when comparing it with other sandboxing approaches. A practical evaluation should check the enforcement layer, default permission model, communication mediation and audit trail, model-file verification, supported engines, deployment environments and whether management is self-hosted or delivered through a paid control plane.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
