Research notes and analysis from the Futurum practice areas.
Fernando Montenegro, VP & Practice Lead, Cybersecurity, argues that a kill switch acts late, stops the whole agent, and depends on detection agents can blind, so containment must bound authority per action and tie autonomy to reversibility.
Fernando Montenegro, Vice President & Practice Lead, Cybersecurity & Resilience at Futurum, examines CrowdStrike's acquisition of Fiddler AI and Dreadnode to launch Falcon Observe, extending its platform's ability to test, monitor, and stop AI agent.
Fernando Montenegro, VP at Futurum, analyzes Cribl’s launch of Cribl Detect and StreamAI, and what a pipeline vendor running the detection loop means for buyers choosing to consolidate or compose.
Fernando Montenegro, VP at Futurum, analyzes Palo Alto Networks’ Unit 42 Continuous Frontier AI Defense, weighing gated model access, tier economics, and whether always-on AI testing closes exposures.
Fernando Montenegro, VP at Futurum, analyzes CrowdStrike’s Fal.Con 2026: an ambitious, largely defensible push into its own AI models and agent security, anchored to the endpoint and the promise to stop breaches.
Fernando Montenegro, VP & Practice Lead at Futurum, shares his insights on how offensive AI succeeds because it’s easier to verify than defensive security, shifting the need for vendors to sell proof of exploitability over volume.
