CrowdStrike (NASDAQ: CRWD) has signed a definitive agreement to acquire Dreadnode, a developer of automated artificial intelligence security testing technology, as the cybersecurity company expands its Falcon platform to address vulnerabilities in enterprise AI systems.
The acquisition will add capabilities for continuously testing AI models, applications and agents before deployment, including identifying weaknesses that conventional software security assessments may miss.
- CrowdStrike has signed a definitive agreement to acquire Dreadnode, expanding its AI security capabilities across models, applications and autonomous agents.
- Dreadnode’s technology uses more than 100 AI attack strategies, including jailbreaks and prompt injection, to identify vulnerabilities before systems enter production.
- Automated AI Security Testing is scheduled for general availability in Q4 FY27, with the acquisition expected to close during the same quarter.
CrowdStrike plans to integrate Dreadnode’s technology into its Falcon platform, extending its AI Detection and Response (AIDR) offering into the development and testing stages of enterprise AI systems.
Enterprises are deploying AI faster than they can prove it is secure. A model can be jailbroken, an agent can be manipulated, and most organizations have no way to find those flaws until the AI is already in production.
George Kurtz, CEO and founder of CrowdStrike
“With Dreadnode, CrowdStrike will give every enterprise the power to trust AI and deploy faster, with confidence.”
Dreadnode brings automated AI security testing to CrowdStrike Falcon
Traditional security testing generally examines software with predictable behavior, whereas AI models and agents produce probabilistic outputs and can respond differently to similar inputs.
This introduces vulnerabilities such as prompt injection, jailbreaks and manipulation of agent reasoning processes that conventional testing methods were not designed to detect.
Dreadnode’s technology addresses these risks by running automated attack simulations against AI systems before they reach production.
Users can describe testing requirements in plain language, after which the platform executes more than 100 known AI attack strategies against models, applications, agent reasoning loops and Model Context Protocol (MCP) servers.
The tests include jailbreak attempts, prompt injection, multi-turn attacks and techniques specifically designed to exploit autonomous AI agents.
Dreadnode also examines AI models, model repositories and related components for security vulnerabilities during development.
The platform ranks identified weaknesses by severity and maps findings to established security frameworks, including the OWASP Top 10 for LLMs, OWASP Agentic Security (ASI01-ASI10), MITRE ATLAS and the NIST AI Risk Management Framework.
Exportable reports allow security teams to assess findings and make deployment decisions based on identified risks.
Human operators can review and reclassify results, adjust severity ratings and document their reasoning. Dashboard metrics are automatically recalculated when assessments change.
CrowdStrike expects Automated AI Security Testing to become generally available in the fourth quarter of fiscal 2027.
CrowdStrike to integrate Dreadnode with Falcon Guardian and SafeMind
The acquisition will connect Dreadnode’s pre-deployment testing capabilities with CrowdStrike’s existing AI security products.
Findings generated during security testing will feed into Falcon Guardian as policy recommendations, allowing weaknesses discovered before deployment to inform protections applied while AI systems are running.
Testing data will also support CrowdStrike’s SafeMind models, which are used to evaluate the security and cyber capabilities of advanced AI systems.
Dreadnode’s researchers will join CrowdStrike’s Cyber Superintelligence Lab, expanding its research capabilities in AI vulnerabilities, attack techniques and model security.
The integration is intended to connect security testing during AI development with runtime detection and response, rather than treating the two activities as separate processes.
AI fails in ways traditional testing was never designed to catch, and the people best equipped to find those failures think like attackers. We founded Dreadnode to expose those weaknesses before AI reaches production.
Brad Palm, CEO of Dreadnode
“As part of CrowdStrike, we will bring Automated AI Security Testing to every enterprise, helping teams confidently find and fix weaknesses.”
CrowdStrike acquisition expected to close in Q4 FY27
CrowdStrike said the proposed purchase price will be paid predominantly in cash, with a portion delivered in company stock subject to vesting conditions. The company did not disclose the transaction’s financial value.
The acquisition is expected to close during CrowdStrike’s fourth quarter of fiscal 2027, subject to customary closing conditions.
Dreadnode will extend the Falcon platform’s existing security capabilities, which cover endpoints, cloud workloads, identity and data protection.
CrowdStrike’s cloud-native platform combines real-time attack indicators, threat intelligence, adversary techniques and enterprise telemetry to support threat detection, automated protection, remediation, threat hunting and vulnerability prioritization.
The planned addition of automated AI security testing will extend those capabilities to weaknesses in AI models and agent behavior before deployment, with findings also contributing to runtime protection through Falcon Guardian.
