OpenClaw, an AI agent tool that automates business tasks by accessing email and calendar systems, faces widespread adoption barriers due to severe security vulnerabilities. In response, a consortium including OpenAI and Red Hat is developing OpenClaw Enterprise (OCE), an enterprise-grade version designed to address cybersecurity concerns and governance requirements that currently prompt most organisations to ban agentic platforms entirely. The project aims to enable safe deployment of powerful agents in business environments without compromising their capabilities.

The original OpenClaw tool was created by developer Peter Steinberger and gained initial acclaim before security analysts identified critical flaws. Gartner classified it as an "unacceptable cybersecurity risk" for business users, whilst China's Computer Emergency Response Team warned of "extremely weak default security configurations." OCE introduces multi-tenancy, hard security boundaries, standardised agent primitives, and comprehensive governance and auditability across the agent lifecycle. Red Hat has positioned OCE analogously to Kubernetes, suggesting the project could transform enterprise adoption of agentic AI systems if it achieves similar industry acceptance.