nFox open-sources AI sandbox Blockhouse for enterprise LLMs

Artificial intelligence governance startup nFox Lab has open-sourced Blockhouse, a security sandbox designed to restrict what AI workloads can access and execute. The company, which emerged from stealth this week, also introduced a platform for managing large language models.

Blockhouse acts as a runtime guardrail, limiting model access to approved data sources and compute resources. The open-source release follows a pattern seen in other recent AI governance tools, such as Comp AI’s compliance automation platform, which raised $34 million in September. The timing aligns with broader industry trends, including AWS’s October open-sourcing of a 2-billion-parameter decision model and Mistral AI’s €3 billion funding round last month.

The focus on security sandboxes may reflect growing interest in tools that address compliance or risk management for AI workloads. However, it remains to be seen how Blockhouse will stand out in a landscape that already includes open-source security frameworks.

Sources: siliconangle.com