Meta and Tech Firms Restrict OpenClaw Over Security Fears

Major technology companies, led by Meta, are moving to restrict the use of the OpenClaw AI agent within their internal development environments and on their platforms. The decision comes in response to mounting security concerns from experts about the agent's unpredictable nature and potential for exploitation. OpenClaw, while demonstrating impressive autonomous capabilities, has been flagged for its lack of built-in safety constraints and its tendency to perform actions in unexpected ways. Security researchers warn that its power could be easily hijacked through prompt injection or other attacks to conduct harmful activities, such as data theft or system compromise. These corporate restrictions reflect a growing caution in the tech industry regarding the rapid deployment of highly capable, open-source AI agents. The move signals a preference for more controlled, auditable AI tools and underscores the industry's recognition that capability must be balanced with robust safety and secu