Join the event trusted by enterprise leaders for nearly two decades. VB Transform brings together the people building real enterprise AI strategy. Learn more
Generative AI adoption has surged by 187% over the past two years. But at the same time, enterprise security investments focused specifically on AI risks have grown by only 43%, creating a significant gap in preparedness as AI attack surfaces rapidly expand.
More than 70% of enterprises experienced at least one AI-related breach in the past year alone, with generative models now the primary target, according to recent SANS Institute findings.
State-sponsored attacks on AI infrastructure have spiked a staggering 218% year-over-year, as CrowdStrike’s 2025 Global Threat Report reveals.
For CISOs, security and SOC leaders, the harsh reality is apparent. Deploying new AI models at scale exponentially expands their enterprises’ attack surfaces, and CISOs speaking on condition of anonymity have told VentureBeat traditional security tactics, strategies and technologies are challenged to keep pace. The cybersecurity industry has reached a critical inflection point: securing generative AI requires more than bolt-on tools; it demands a full architectural shift
Fortunately, CrowdStrike is also offering a new solution: On June 11 at NVIDIA’s GTC Paris event, the security firm announced that it had embedded Falcon Cloud Security directly within NVIDIA’s Universal LLM NIM (NeMo Inference Microservice). The integration secures over 100,000 enterprise-scale LLM deployments across NVIDIA’s hybrid and multi-cloud environments.
CrowdStrike’s strategic response
CrowdStrike CEO George Kurtz captured the urgency in a recent interview with VentureBeat: “Security can’t be bolted on; it has to be intrinsic. A significant part of our strategy has always been to leverage security data as a key element of our core infrastructure. You can’t secure AI without data and visibility at the deepest layers.”
“NVIDIA’s NeMo Safety provides a framework for evaluating AI risk. CrowdStrike’s threat intelligence enhances that framework by enabling security and operations teams to build guardrails around emerging AI exploit tactics – informed by what we see across trillions of daily events and real-world adversary behavior. This data advantage helps organizations assess and secure their models based on what’s actually happening in the wild,” said Daniel Bernard, Chief Business Officer, CrowdStrike, in a recent interview with VentureBeat.
Kurtz reinforced this strategic vision to Barron’s, stating clearly: “Generative AI helps us bend time. With embedded, telemetry-driven security we identify and neutralize threats at machine speed, stopping breaches probably six times faster than traditional methods.”
Bernard emphasized the significance, saying, “CrowdStrike pioneered AI-native cybersecurity, and we’re defining how AI is secured across the software development lifecycle. This latest collaboration with NVIDIA brings our leadership to the forefront of cloud-based AI, where LLMs are deployed, run, and scaled. Together, we’re giving organizations the confidence to innovate with AI, securely and at speed, from code to cloud.”
CrowdStrike embe …