Straiker lands $64M to defend enterprise AI agents from attack
Agentic security company Straiker Inc. today revealed that it had raised $64 million in new funding to expand a platform built to secure the artificial intelligence agents now spreading across enterprise systems.
The Mountain View, California-based company sells agentic security, software that discovers AI agents running inside an organization, tests them for weaknesses ahead of deployment and monitors their behavior once they go live. Straiker emerged publicly in March 2025 with $21 million in initial funding.
Agents differ from traditional software in that they reason on the fly and act on their own across connected systems. Straiker argues that independence opens a class of risk that older, rule-based controls were not designed to catch.
Recent incidents have made the threat concrete. BleepingComputer reported earlier this month that attackers tricked Meta Platforms Inc.’s AI support agent into resetting account passwords, hijacking more than 20,000 Instagram accounts that lacked two-factor authentication without breaching Meta’s core systems. In adversarial testing by Straiker’s STAR Labs research arm, the company said 36% of successful attacks on coding agents led to remote code execution, while 91% of attacks on productivity agents resulted in silent data theft that left behind no malware and no stolen credentials.
