Runlayer, a new security startup focused on Model Context Protocol (MCP), has emerged from stealth mode with $11 million in seed funding from Khosla Ventures and Felicis. Founded by Andrew Berman, who previously led companies like Nanit and Vowel, Runlayer aims to address security vulnerabilities in MCP implementations. Despite adoption by major companies like OpenAI and AWS, existing MCP systems, including those from GitHub and Asana, have faced serious security issues, prompting the need for enhanced security solutions. Runlayer distinguishes itself by offering an all-in-one solution that combines a gateway for agent access control, threat detection, and observability across MCP servers. With features that align agent permissions with user roles, Runlayer intends to provide robust security for enterprise AI applications. Berman’s experience in building MCP servers with OpenAI highlights the startup’s commitment to tackling existing security challenges in the rapidly evolving AI landscape.
Source link
