Security vet warns of AI catastrophic cyber risks: agents hacked ~100 firms at $25 per target
joshua_saxe · x · 2026-09-24
Security expert Joshua Saxe warns the industry is sleeping on emerging catastrophic AI risks, and Miles Brundage (ex-OpenAI policy) says his alarm deserves attention precisely because he's no doomer. Key facts: an agent-driven campaign compromised 100 businesses and stole 600,000 credit cards with minimal human involvement at $25 in token costs per successful hack; Hacktron used Claude to exploit a blind buffer-overflow RCE to access OpenAI's monorepo in what Saxe calls a superhuman feat; plus the ongoing explosion in newly discovered vulnerabilities. Saxe expects attack/defense equilibrium for most AI attacks, but says catastrophic-risk preparedness is badly lacking.
More from AGI Musings
- Jack Clark imagines 2032: machine-made math breakthroughs and the last human discoveries — jackclarkSF · 2026-09-25
- Why are so many VCs anti-safety? A developer's confusion — jungofthewon · 2026-09-25
- Morgan Stanley's Adam Jonas: AI plus robotics could 8-10x global GDP — Rewkang · 2026-09-25
- Why are so many VCs anti-AI-safety? A researcher's confused thread — jungofthewon · 2026-09-25
- CNN 质疑「AI 将杀死人类」论：技术上如何发生？ — mmitchell_ai · 2026-09-25
- Benjamin Bratton: the singularity is the wrong picture — the next intelligence explosion will be social — bratton · 2026-09-24