CrowdStrike is doubling down on artificial intelligence security. The cybersecurity vendor announced that it has reached an agreement to acquire Pangea, a startup focused on providing guardrails for generative AI–powered applications.
The deal, valued at approximately $260 million, according to The Wall Street Journal, comes at a time when AI adoption across enterprises is accelerating rapidly, while new attack methods are emerging to match the pace.
Securing the AI lifecycle: CrowdStrike’s next frontier
CrowdStrike CEO George Kurtz announced the deal on stage at the Fal.Con 2025 conference in Las Vegas, framing it as the next step in the company’s journey. Just as CrowdStrike set the standard for endpoint protection with EDR, he said, the goal now is to bring that same approach to AI.
“AI is rewriting the enterprise attack surface at breakneck speed. Each prompt becomes an entry point for the adversary,” said Kurtz. “With Pangea, CrowdStrike will secure the entire AI lifecycle, detecting risks, enforcing safeguards, and ensuring compliance, so our customers can confidently build, deploy, and scale AI without risk.”
By bringing Pangea into the Falcon platform, CrowdStrike is rolling out what it calls AI Detection and Response (AIDR). The idea is simple: take the same approach that made EDR the standard for endpoint security and apply it to AI, which covers everything from development to day-to-day use.
Guardrails against prompt-based threats
Pangea’s technology focuses on protecting AI agents and workflows at every layer, from browsers and applications to cloud environments. That includes stopping threats like prompt injection, model jailbreaks, and other such manipulations. Pangea says its defenses can catch up to 99% of bad prompts with sub-30 millisecond response times.
Oliver Friedrichs, Pangea’s founder and CEO, underscored the startup’s mission: “Pangea was founded to make AI adoption safe and secure, giving enterprises the visibility and guardrails to embrace AI with confidence. By joining CrowdStrike, we will be able to deliver this vision on a global scale, unifying AI security with the Falcon platform and creating the industry’s first complete AI Detection and Response platform.”
Building for the next wave of AI
CrowdStrike definitely has its eye on the AI prize. The company has already partnered with Nvidia and AWS on AI security initiatives and recently announced plans to acquire Onum, a startup specializing in data pipeline management, to enhance its Falcon Next-Gen SIEM.
Kurtz framed the stakes clearly during Fal.Con: “But we can’t be part of the change unless we secure it. Securing AI [is] going to be a big part of the future growth opportunity for us and our partners.”
By extending Falcon to cover prompts, responses, and agent communications, CrowdStrike is betting it can do for AI security what it once did for endpoint protection: establish the industry standard.
The push into AI security isn’t unique to CrowdStrike. Just this summer, Snyk bought Invariant Labs to bolster its AI Trust Platform, adding research expertise around agentic attack vectors and runtime detection. Both moves highlight how vendors are racing to put guardrails in place as enterprises scale up their use of AI. Read more here.