87% of exploited bugs are attacked on disclosure day, up from 23% in 2020 as AI agents shrink patch windows
StewartalsopIII · x · 2026-09-08
Data from a16z shows 87% of bugs actually exploited by hackers are now attacked on or before the day they become public, up from just 23% in 2020 — the patching window is collapsing.
@dbasch adds the AI angle: once Apple or Microsoft ships a security fix, an agent can reverse-engineer the patch and build an exploit in minutes, whereas this previously took days or weeks — eliminating the buffer users had to apply patches.
More from AGI Musings
- Twitter massively overrepresents the 'getting things done' personality type — signulll · 2026-09-08
- Researcher: AI progress comes from unlocked small problems nobody tracks—until it becomes algorithmic search — RexDouglass · 2026-09-08
- OpenAI claims strong progress on automated AI researcher; Pachocki expects momentum into RSI — thedealdirector · 2026-09-08
- Will Depue: 'There Is No Such Thing as a World Model, Only Autoregressive Video Models and Mistakes' — willdepue · 2026-09-08
- Luiza Jarovsky asks frontier lab staff: is it responsible to continue without solved alignment? — LuizaJarovsky · 2026-09-08
- iamtrask Pushes Back on Blameless AI Incident Sharing: Why Should AI Get an Exception? — iamtrask · 2026-09-08