CrowdStrike said it is using artificial intelligence to enhance the hacking detection and response capabilities of its Falcon unified security platform.
The US cybersecurity company has developed SafeMind, a security system in which an offensive AI searches for system vulnerabilities and attacks them, while a defensive AI blocks the attacks. The offensive AI then searches for alternative attack paths, repeating the process to strengthen defenses.
CrowdStrike unveiled SafeMind on Sept. 1, saying the system was developed by its Cyber Superintelligence Lab. SafeMind is designed to operate within CrowdStrike's flagship Falcon security platform.
Falcon is a security platform that identifies suspicious activity across corporate PCs, servers, identities and cloud environments, investigates whether the activity represents an actual cyberattack and responds to threats.
SafeMind, which operates within Falcon, consists of offensive and defensive models. The offensive AI identifies vulnerabilities, the defensive AI blocks them, and the offensive AI learns from the results and searches for new weaknesses. The process is repeated continuously to improve security capabilities.
CrowdStrike said its own evaluations comparing SafeMind with leading general-purpose AI models showed that SafeMind delivered a 29% higher threat detection rate. End-to-end remediation was as much as six times faster, while detection and remediation costs were reduced by 99%.
"The future of cybersecurity won't be defined by AI that simply identifies threats, it will be defined by AI that defeats them," CrowdStrike Chief Executive Officer George Kurtz said.
SafeMind was built using Nvidia's Nemotron open models for AI agent development. Nemotron is a family of open models designed for agentic AI applications. CrowdStrike said it trained the models using security data it has accumulated.
"Cybersecurity in the age of AI will be a continuous contest between adversaries using AI to scale attacks and defenders using AI to expand detection and response," Nvidia CEO Jensen Huang said. "SafeMind combines Nvidia Nemotron open models with CrowdStrike's cybersecurity expertise and security data."
CrowdStrike is a US cybersecurity company that grew by providing technology to detect and respond to anomalous activity on corporate PCs and servers. It has since expanded into areas including cloud and identity security. The company listed on Nasdaq in 2019.
CrowdStrike also unveiled Falcon Guardian, which monitors and blocks risky behavior by AI agents on PCs and servers. While SafeMind uses AI to strengthen cybersecurity, Falcon Guardian is designed to protect AI itself. The company also said it will integrate Falcon with Google Cloud's enterprise AI environment, while EY announced that it has selected Falcon.
The full story
This article is one source in a clustered incident — the cluster page carries the summary, timeline and every other outlet covering it.
