Analysis and insights on AI-driven threats, Zero Trust architecture, and the future of network security from the OpenNHP team.
A month of public testing, a hackathon full of sharp people trying to break it, and OpenNHP is still standing. Slides and results from the IETF 126 Hackathon session are now available.
Can you discover or access infrastructure that's designed to be invisible? OpenNHP is joining the IETF 126 Hackathon with a simple challenge: try to break it. We welcome security researchers, cryptographers, protocol designers, and anyone who enjoys attacking assumptions.
AI is discovering vulnerabilities faster than humans can fix them. Project Glasswing found over 10,000 high- and critical-severity vulnerabilities in one month, confirming that the era of visible infrastructure is ending.
In April 2026, an AI model completed a 32-step enterprise attack end-to-end for the first time. The UK AI Safety Institute, NCSC, and multiple 2026 studies now converge on one conclusion: offensive AI capability is doubling every 4 months. Here's why network hiding is the only rational defense.
Two product launches this week—PentAGI and Claude Code Security—mark the point where AI attack capabilities became freely available to anyone. Taken together with two years of academic research, they confirm a simple equation: if you are visible, you are vulnerable.