CrowdStrike & NVIDIA Team Up to Combat Machine-Speed Cyberattacks
In the relentless battle against cyber threats, security operations center (SOC) leaders often find themselves overwhelmed, struggling to keep pace with attacks that are increasingly automated and sophisticated. Recognizing this challenge, CrowdStrike and NVIDIA have joined forces to flip the script, introducing a new era of proactive defense.
A New Arms Race in Cybersecurity
The collaboration between CrowdStrike and NVIDIA marks a significant shift in cybersecurity. By combining the strengths of open-source technology with agentic AI, they are empowering security teams to strike back at attackers with unprecedented speed and precision. This initiative leverages the power of autonomous agents, including Charlotte AI and NVIDIA Nemotron models, to transform how organizations approach threat detection and response.
Key Components of the Agentic Ecosystem
The core of this collaboration is built on a robust agentic ecosystem, including:
- Charlotte AI AgentWorks
- NVIDIA Nemotron open models
- NVIDIA NeMo Data Designer for synthetic data
- NVIDIA Nemo Agent Toolkit
- NVIDIA NIM microservices
According to Bryan Catanzaro, vice president of Applied Deep Learning Research at NVIDIA, this partnership redefines security operations by enabling analysts to build and deploy specialized AI agents at scale, utilizing trusted, enterprise-grade security with Nemotron models. The goal is to enable autonomous agents to learn quickly, thereby reducing risks, threats, and false positives, and alleviating the burden on SOC teams.
Transforming Analyst Expertise into Machine-Scale Datasets
A key differentiator of this partnership is the ability of AI agents to continually aggregate telemetry data, including insights from CrowdStrike Falcon Complete Managed Detection and Response (MDR) analysts. Dan Bernard, CrowdStrike’s Chief Business Officer, explained that they are turning the intelligence, data, and experience of their Falcon Complete analysts into datasets, AI models, and agents. This allows customers to benefit from these agents at scale.
Organizations will be able to have their autonomous agents continually learn by training on the datasets from Falcon Complete, the world’s largest MDR service handling millions of triage decisions monthly. CrowdStrike’s experience in AI detection triage has already yielded significant results, such as the Charlotte AI Detection Triage, which integrates into existing security workflows, automates alert assessment with over 98% accuracy, and reduces manual triage time by more than 40 hours per week.
Elia Zaitsev, CrowdStrike’s chief technology officer, emphasized the crucial role of the Falcon Complete team in achieving such performance, highlighting the value of their high-quality, human-annotated dataset.
Embracing Open Source for Enhanced Security
NVIDIA’s Nemotron open models are designed to address the critical need for transparency and control in AI adoption, especially in regulated environments. Justin Boitano, Vice President, Enterprise and Edge Computing at NVIDIA, noted that open models are essential for building specialized domain knowledge, allowing organizations to maintain data privacy and security. John Morello, CTO and co-founder of Gutsy, emphasized that open-source models offer more transparency, better assurances of data privacy, and greater integration options.
Cisco’s DJ Sampath, senior vice president of Cisco’s AI software and platform group, highlighted the industry-wide imperative for open-source security models, emphasizing the need to empower defenders with robust models. NVIDIA’s commitment to transparency extends to sensitive aspects of AI development, as demonstrated by their response to concerns about DeepSeek R1’s training data. Itamar Sher, CEO of Seal Security, underscored the importance of transparency in open-source models for practitioners managing security at scale.
Bringing Intelligence to the Edge
A key goal of this partnership is to bring intelligence closer to where data is and decisions are made, which is especially critical for government agencies. NVIDIA and CrowdStrike have already laid the groundwork, including STIG hardening, FIPS encryption, and air-gap compatibility, to facilitate the adoption of open models on higher-side networks. Boitano emphasized the urgency of having AI defense running in your estate to protect against the speed of AI.
About the Author
