Evoke Security Raises $4M to Secure AI Agents
Evoke Security has raised a $4 million pre-seed round to build a security platform for autonomous AI agents. The company aims to inventory, monitor, and prevent unauthorized actions from what it calls the "agentic workforce," addressing risks such as data leakage and protocol abuse in automated systems.
- The $4 million pre-seed funding round was led by Crosspoint Capital Partners, with participation from Red Cell Partners. - Evoke Security was recently selected to participate in the 2026 CrowdStrike, AWS, and NVIDIA Cybersecurity Startup Accelerator, an elite eight-week program. - The company's platform addresses risks like excessive permissions, agent hijacking, and tool misuse, which are common vulnerabilities in autonomous AI systems. - Evoke Security's technology is designed to deploy in minutes and provides inventory and governance of AI agents, AI security posture management, and threat modeling to identify over-permissioned agents. - The platform uses a lightweight endpoint sensor, a browser extension for SaaS agents, and an SDK and proxy for production agents to ensure consistent policy and telemetry. - Co-founder and CEO Jason Rebholz was formerly the CISO at Corvus Insurance, and co-founder and CTO is Jeffrey Chan. - The company aims to tackle the security challenge of AI agents accessing sensitive enterprise data, such as customer information, source code, and production systems, often without monitoring from traditional security tools. - Key threats the platform is designed to mitigate include prompt injection, identity spoofing, and data leakage through model context, which are specific attack vectors for AI agents.
