CrowdStrike has unveiled a new artificial intelligence system designed to help cybersecurity teams detect and respond to attacks, as companies race to deploy AI tools capable of acting autonomously.
The system, called SafeMind, combines purpose-built offensive and defensive AI models with software “harnesses” that allow the models to operate in a continuous loop, CrowdStrike said at its Fal.Con 2026 conference in Las Vegas.
The company explains SafeMind was built using NVIDIA’s open Nemotron models and trained with data from CrowdStrike’s Falcon cybersecurity platform, including threat intelligence, incident response data and telemetry from its sensors.
SafeMind includes two models: Red Tempest, designed to simulate advanced attacks, and Blue Solano, designed to identify and defend against threats.
CrowdStrike says its internal evaluations found that SafeMind’s defensive model delivered a 29% higher detection rate, six times faster end-to-end remediation and 99% lower costs for detection and remediation compared with leading frontier models and open-source baselines. Those figures are company evaluations and were not independently verified.
“The future of cybersecurity won’t be defined by AI that simply identifies threats, it will be defined by AI that defeats them,” said George Kurtz, CEO and founder of CrowdStrike. “SafeMind brings offensive and defensive models together in a system trained on CrowdStrike’s unique cyber data. It finds weaknesses, strengthens protection, and gets smarter with every cycle, advancing our mission to stop breaches at machine speed.”
NVIDIA CEO Jensen Huang said the development reflected a shift in which both attackers and defenders are increasingly using AI to automate cybersecurity operations.
“We’re at an inflection point in cybersecurity,” Jensen Huang told a sold-out crowd at CrowdStrike’s Fal.Con 2026 in Las Vegas Tuesday. Attacks are now automated. Defense has to be, too.
“This is the beginning of a new age of cybersecurity,” Huang told the crowd of 10,000 security professionals. “On the one hand, the adversaries are going to be more armed than ever. On the other hand, all of you are going to be more armed than ever.”
SafeMind is designed to allow an offensive model to identify an attack path while a defensive model attempts to close it, with the results feeding back into the system to improve future responses.
NVIDIA said its evaluation used an isolated cyber environment modeled on its own accelerated computing infrastructure. The test linked offensive and defensive agents in a continuous loop, with the defensive system generating and validating detections based on telemetry.
CrowdStrike said the system uses NVIDIA Nemotron 3 Ultra for defensive orchestration and a post-trained Nemotron 3 Super model for detection generation and repair. The company said it trained the models on its own cybersecurity data.
“Together with NVIDIA, we built cybersecurity’s first complete agentic system for cybersecurity, including the first frontier models and harness purpose-built for defenders,” Kurtz said. “This isn’t a copilot baked into someone else’s intelligence. It’s not a chatbot with a security skin. It is a frontier-class model built and trained by CrowdStrike on our data in partnership with NVIDIA.”
Huang said the combination of models and software harnesses was important because the latter can turn a general-purpose AI model into a system capable of taking actions in a particular environment.
“The harness is essentially the exoskeleton of the large language model,” Huang said. “The large language model is the brain. The exoskeleton turns it into an agent — and this exoskeleton doesn’t have to be the same shape and capability for every domain.”
CrowdStrike also announced Falcon IQ, a system that uses more than 50 AI agents to automate cybersecurity assessment, prioritization and remediation workflows. It is built on CrowdStrike’s Falcon platform and Charlotte AI AgentWorks, the company’s no-code platform for developing AI agents.
Falcon IQ is part of Project QuiltWorks, CrowdStrike’s initiative to bring together cybersecurity companies, systems integrators, cloud providers and insurers around securing AI systems.
“Falcon IQ changes the economics of securing frontier AI risk,” said Daniel Bernard, chief business officer at CrowdStrike. “Agents do the heavy lifting, partners scale delivery, and organizations gain protection in a fraction of the time and cost. QuiltWorks proved the model, Falcon IQ advances it from manual to the speed of AI across the Falcon platform.”
The announcements come as businesses deploy AI agents that can access enterprise systems and execute tasks with limited human intervention, creating a new security challenge for companies. CrowdStrike has said AI-enabled attacks rose 89% in the past year, while the fastest observed eCrime breakout time fell to 27 seconds.
CrowdStrike also announced Falcon Guardian, an AI detection and response system intended to monitor AI agents at runtime and identify threats including prompt injection, sensitive-data leakage and malicious AI activity.
“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. Cyber defense will be among the most compute-intensive applications of AI,” Huang said. “SafeMind combines NVIDIA Nemotron open models with CrowdStrike’s deep cybersecurity expertise, trusted security data, purpose-built agent harnesses, rigorous evaluations, and safeguards – creating a frontier agentic cybersecurity stack designed to operate at machine speed. Powered by NVIDIA accelerated computing, SafeMind makes AI a force multiplier for defenders.”
CrowdStrike has also established a Cyber Superintelligence Lab focused on developing AI systems for cyber defense, bringing together its AI researchers, offensive security specialists and incident responders.
“Our models are the start of a new chapter for cyberdefense,” said Dr. Bartley Richardson, chief AI and autonomous systems officer at CrowdStrike. “With the models and harnesses together in a co-evolving agentic system, defenders can now act at machine speed. This is the foundation for the next decade of AI security, and CrowdStrike is the only company that owns the entire stack, from sensor to harness to model.”
The companies said SafeMind can also work with other frontier and open-source AI models, allowing customers to use different models while retaining CrowdStrike’s cybersecurity harnesses.
“The real test of AI is what it can do in production, at scale, when the stakes are highest. Few environments put that to the test more than cybersecurity,” said Michael Intrator, co-founder and chief executive officer, CoreWeave. “We’re proud to power SafeMind across training and inference as CrowdStrike puts specialized AI to work against real-world threats.”
Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.
Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.
Got a story to share? Contact Us to reach a global audience with Impact Newswire.
Faustine Ngila is the AI Editor at Impact Newswire, based in Nairobi, Kenya. He is an award-winning journalist specializing in artificial intelligence, blockchain, and emerging technologies.
He previously worked as a global technology reporter at Quartz in New York and Digital Frontier in London, where he covered innovation, startups, and the global digital economy.
With years of experience reporting on cutting-edge technologies, Faustine focuses on AI developments, industry trends, and the impact of technology on society.
Discover more from Impact Newswire
Subscribe to get the latest posts sent to your email.
