NVIDIA and CrowdStrike Unveil SafeMind: Agentic Cybersecurity With Red-Blue Coevolution
NVIDIA Blog · rss · 2026-09-02
At CrowdStrike's Fal.Con 2026, Jensen Huang and CEO George Kurtz launched CrowdStrike SafeMind, an agentic cybersecurity system built on NVIDIA Nemotron. Context: AI-enabled attacks rose 89% last year and the fastest eCrime breakout time is now 27 seconds — "human-speed response isn't defense, it's documentation."
Key points:
- SafeMind uses Nemotron 3 Ultra to orchestrate the defensive agent harness, with a fine-tuned Nemotron 3 Super powering rule generation, post-trained on CrowdStrike's 15 years of threat data. Internal evals show its Blue Solano model beats leading frontier models in accuracy at 99% lower cost.
- It runs a continuous red-blue coevolution loop: red-team agents find exploits in a digital twin of NVIDIA's network while blue-team agents generate and validate detections via Falcon sensors. Huang says the framework extends to robotics and edge computing.
- CrowdStrike also announced Falcon IQ (50+ coordinated agents automating remediation workflows) and expanded its Guardian AI safety solution. Open Nemotron lets CrowdStrike post-train without exposing data — something closed frontier models can't offer.
More from Companies & People
- 562 AI PM job postings split into 4 roles — Model PM tops pay at $345K median — aakashgupta · 2026-10-18
- SGLang Summit sponsors grow from 11 to 36 in one week; Lip-Bu Tan, Lilian Weng to speak — BanghuaZ · 2026-10-18
- Yacine MTB: Users will remember how Anthropic treated them once intelligence proliferates — yacineMTB · 2026-10-18
- Periodic's CoLM talk: when does training your own model beat frontier + harness? — heghbalz · 2026-10-18
- After GHC Cancellation, GitHub and Cloudflare Offer 175 Free Conference Passes — michellechen · 2026-10-18
- Ex-Apple Siri Engineer Builds 7 iOS Apps Solo, Hits $6.2K MRR — jdluk87 · 2026-10-18