CrowdStrike & NVIDIA launch AI collaboration in security

CrowdStrike has announced new agentic AI innovations integrated with NVIDIA AI software to enhance cybersecurity measures.

The advancements, which employ NVIDIA AI software tools including NVIDIA NIM microservices, promise substantial improvements for Security Operation Centres (SOCs).

These developments enable CrowdStrike's Charlotte AI Detection Triage to perform at twice the speed compared to its initial launch, all while utilising 50% fewer GPUs, thus optimising resource use.

CrowdStrike and NVIDIA's collaboration is pushing the boundaries of cybersecurity by employing advanced reasoning models.

These models are intended to enhance intelligent automation, enable real-time decision making, and improve precision in threat response efforts.

The 2025 CrowdStrike Global Threat Report cites that adversaries can sometimes infiltrate systems in as little as 51 seconds, rendering traditional security tools insufficient. Overwhelming analysts with numerous alerts, the effectiveness of these tools has diminished, leaving organisations exposed to risks. Through the use of full-stack NVIDIA AI-accelerated computing, CrowdStrike intends to provide its clients with the necessary automation and intelligence to keep pace with modern threats.

The collaboration also explores NVIDIA's Llama Nemotron reasoning models, which aim to further enhance detection accuracy.

This technology is designed to empower security teams, ensuring a rapid and precise response to potential attacks.

Daniel Bernard, Chief Business Officer at CrowdStrike, expressed the potential of agentic AI in cybersecurity, stating, "The future of cybersecurity is agentic AI—where advanced reasoning models power intelligent automation to work seamlessly with human analysts to stop breaches faster than ever." He emphasised the combination of CrowdStrike's Falcon platform with NVIDIA AI as a step towards the next evolution in security operations.

The enhancements in AI-powered security operations are significant, improving triage processes. With Charlotte AI Detection Triage leveraging NVIDIA's software, detection processing has been accelerated, reducing operational costs.

Additionally, efforts to apply reasoning models are ongoing.

These models are expected to minimise false positives, allowing security teams to focus their attention on credible threats with heightened precision.

Continuing the trend towards automated security measures, CrowdStrike and NVIDIA are developing AI agents to support analysts in triage, response, and analysis on a large scale, paving the way for future security automation.

Kari Briski, Vice President of Generative AI Software for Enterprise at NVIDIA, described the collaboration's ambitions: "Enterprises need AI that doesn't just react but anticipates—intelligent systems with advanced reasoning that adapt, decide and respond in real time."

She highlighted the capability of the collaboration to deliver agentic AI to bolster security operations, enhancing resilience in a dynamic digital environment.