Open-Source AI Governance Framework Launches to Secure Multi-Agent Systems

Levels Of Self, led by Arthur Palyan, has launched an open-source framework for AI governance called the Nervous System Framework. The tool is designed to manage and secure multi-agent AI systems across multiple messaging platforms while ensuring strict compliance with emerging regulations.

The Nervous System Framework oversees 13 autonomous AI agents operating on five major platforms: Telegram, WhatsApp, Instagram, Facebook Messenger, and web interfaces. It functions as a Model Context Protocol (MCP) server, providing features such as configuration drift detection and automated compliance checks.

Since its deployment, the system has already identified and prevented over 99 governance violations in production environments. Companies using the framework report cost reductions of 40–60%, thanks to its ability to automate repetitive tasks and streamline operations. Arthur Palyan highlights that the framework's governance layer is a key differentiator, particularly as regulations like the U.S. Executive Order 14110 and the EU AI Act demand auditable AI systems. The launch aligns with rising industry interest in multi-agent AI orchestration, further validated by NVIDIA's recent release of NemoClaw—a similar multi-agent framework that supports Levels Of Self's architectural approach.

The Nervous System Framework provides a scalable solution for businesses needing secure, compliant AI operations. Its ability to cut costs while maintaining strict governance positions it as a practical tool for industries adopting multi-agent systems. The framework's open-source nature also encourages broader collaboration and refinement within the AI community.