Local AI offers organizations greater privacy, cost control, and data ownership, but running models locally does not eliminate security risks. Two llama-server vulnerabilities received CVSS scores of 9.2 , highlighting the potential risk to organizations exposing local AI inference services to untrusted traffic. , highlighting the potential risk to organizations exposing local AI inference services to untrusted traffic. Local inference instead performs these operations on infrastructure controlled by the user or organization. How concurrency flaws create memory safety risks in local AISeveral findings demonstrate how concurrency can transform ordinary memory-management mistakes into security vulnerabilities.