AIToday
PR TIMES (AI)Published: Nov 13, 2026, 06:01 JST

CrowdStrike to acquire Dreadnode for pre-deployment AI security testing

Send on LINE
CrowdStrike to acquire Dreadnode for pre-deployment AI security testing

3 Key Points

  1. What happened

    CrowdStrike signed a definitive agreement to acquire Dreadnode, whose technology will extend the Falcon platform to test AI models, applications and agents for weaknesses before deployment, with general availability planned for Q4 of fiscal 2027.

  2. Why it matters

    AI models and agents behave probabilistically and can be jailbroken or manipulated by prompt injection, so pre-deployment testing addresses gaps that traditional code-based security testing cannot cover.

  3. What to watch

    General availability of the automated AI security testing is planned for Q4 of fiscal 2027, and the acquisition is expected to close in that same quarter.

WHO IT HITSEnterprise security teams responsible for AI deployments gain a way to test models, agents and applications for jailbreaks and prompt injection before they reach production. AI platform owners integrating third-party models and MCP servers may also need to plan for this type of pre-deployment scanning.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

CrowdStrike's move reflects a gap it sees between how fast companies deploy AI and how slowly they can prove it is safe. Traditional security testing assumes code that behaves the same way every time it runs, but AI models and agents act probabilistically and can be jailbroken or manipulated through prompt injection. Those weaknesses sit in the reasoning loops of models and agents, an area conventional testing does not reach.

Dreadnode's technology is meant to close that gap inside the Falcon platform. It lets teams specify tests in natural language and runs more than 100 known AI attack strategies against enterprise AI models, agent reasoning loops, applications and MCP servers. Findings are ranked by severity and mapped to OWASP Top 10 for LLMs, OWASP Agentic Security (ASI01-ASI10), MITRE ATLAS and NIST AI RMF, with exportable reports so teams can decide whether to deploy.

The deal also folds Dreadnode's team into CrowdStrike's Cyber Superintelligence Lab, and its test data is intended to strengthen CrowdStrike's SafeMind model, while findings feed policy recommendations into Falcon Guardian. CrowdStrike expects the acquisition to close in Q4 of fiscal 2027, the same quarter it plans general availability for the automated AI security testing.

FAQ
What kinds of AI attacks does Dreadnode test for?
Dreadnode runs more than 100 known AI attack strategies, including jailbreaks, prompt injection, multi-turn attacks and agent-specific attacks, against models, agent reasoning loops, applications and MCP servers.
How does Dreadnode report and prioritize findings?
It ranks findings by severity and maps them to frameworks including OWASP Top 10 for LLMs, OWASP Agentic Security (ASI01-ASI10), MITRE ATLAS and NIST AI RMF, and provides exportable reports.
How will CrowdStrike pay for the acquisition?
The purchase price is to be paid mainly in cash, with part delivered as stock subject to vesting conditions.
Send on LINE

Also reported by Top Companies AI

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Next articleBlock opens Builderlab to sell Buzz AI-agent teams