Security · November 6, 2026
CrowdStrike Acquires Fiddler AI and Dreadnode for $500 Million
CrowdStrike has agreed to acquire two artificial intelligence security startups, Fiddler AI and Dreadnode, in a deal reportedly valued at nearly 500 million US dollars. The acquisitions are designed to enhance the Falcon Platform by adding capabilities for AI observability and automated security testing, specifically addressing the risks associated with autonomous AI agents operating in production environments.
The integration of these companies aims to close security gaps in the AI development and usage lifecycle. Fiddler AI, which will operate as Falcon Observe, focuses on post-deployment monitoring. It tracks the decision-making processes and reasoning of AI agents to detect behavioral drift or hallucinations. The system generates audit logs and supports popular frameworks such as LangGraph, Amazon Bedrock, and Google ADK. CrowdStrike CEO George Kurtz described this capability as a black box flight recorder for every AI agent.
Dreadnode provides pre-deployment security testing by running over 100 automated attack simulations, including jailbreak and prompt injection attempts. The company ranks the severity of these threats according to international standards, specifically the OWASP Top 10 for LLMs and the NIST AI Risk Management Framework. This approach addresses the limitations of traditional security tools, which are designed to inspect predictable code but cannot deeply examine the model layer or the reasoning processes of AI systems.
Alongside the acquisitions, CrowdStrike introduced the Falcon Guardian function and a new AI Kill Switch mechanism. This switch allows for the immediate external termination of an agent at the runtime level when anomalies are detected, ensuring that humans retain ultimate control. The company noted that autonomous agents often possess correct access rights and connect to approved tools, yet they may still make incorrect decisions, ignore instructions, or generate false data.
George Kurtz stated that the direction of security is shifting from blocking external attackers to ensuring safe usage. He indicated that an organization's confidence in its ability to control or stop a system promptly will become a primary factor enabling businesses to adopt AI agents more effectively and rapidly.