AI-found vulnerabilities triggered rushed patches that caused outages, NSA tells Congress
ambaonadventure · x · 2026-11-18
Per a Wall Street Journal report, Wall Street execs and the NSA told Congress that recent outages weren't caused by AI itself: AI models surfaced so many vulnerabilities that the rush to patch them before an attack overwhelmed critical systems. Safety researcher Heidy Khlaaf criticized the celebration of this, citing the well-known tension between safety and security — rapid patching in critical infrastructure can itself trigger failures — and calling it reckless behavior driven by AI fear-mongering and poor threat modeling.
More from AGI Musings
- Real AI safety risk is a feudal system by big players, doomer policy doomed to fail — ctjlewis · 2026-11-18
- AI prototypes that look shippable: design teams worry about false signals of completeness — round · 2026-11-18
- Baidu CEO warns AI penetration 'fell short of expectations'; Marcus mocks trillion-dollar IPOs — GaryMarcus · 2026-11-18
- IMF: Circular AI deals top $1 trillion, about two-thirds borrowed — GaryMarcus · 2026-11-18
- Why Silicon Valley now celebrates individuals over companies: influencer culture — dansitu · 2026-11-18
- DeepMind paper argues AGI emerges from millions of specialized agents, not one model — mikeflache · 2026-11-18