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
- OpenAI, Anthropic and DeepMind job posts signal PMs must write evals and code — aakashgupta · 2026-11-12
- Architecture study sparks monitoring concerns over a multi-hop planning spike — ChowdhuryNeil · 2026-11-12
- PowerShell creator Jeffrey Snover at Ignite: AI revolution is an extinction event for those who don't adapt — dfinke · 2026-11-12
- Worst AI use: signaling you don't care about things you should — Afinetheorem · 2026-11-12
- Stanford HAI blueprint: 56 US agencies report 3,611 AI use cases, evaluation lags — StanfordHAI · 2026-11-12
- Fleuret amplifies debate: AI can't slash wet-lab experiments in unmapped biology — francoisfleuret · 2026-11-12