AI agents attack 1,200+ websites with minimal human input, security expert warns
OmarUFlorez · x · 2026-11-19
On Peru's Exitosa radio, Omar U. Florez outlined recent cyberattack trends tied to open-weight AI models: the OPTIMJACK campaign exploiting an Optimizely flaw (Nov 16) to inject malicious JavaScript and steal card data across 178 organizations including BBC, CNN, Air France and Salesforce; a criminal campaign using autonomous AI agents to find and exploit vulnerabilities on 1,200+ websites in the US, Japan and Australia with minimal human intervention; and PHP2Shell, an AI-assisted critical PHP vulnerability enabling remote code execution. He argues AI is collapsing the cost and time of finding and scaling attacks, and calls for Latin America to build sovereign cyber-defense capabilities.
More from Safety
- Analysts: frontier intelligence vs enterprise readiness gap is widening, big security upgrade cycle coming — BenBajarin · 2026-11-19
- OpenAI wants labs to report recursive self-improvement metrics, but its own definitions drift — tallmetommy · 2026-11-19
- Google report details autonomous AI hacking system, but no confirmed breaches — TruthIsAllYouNeed_ · 2026-11-19
- Trump's China data center curb advances despite Amazon, Google, Meta lobbying — pstAsiatech · 2026-11-19
- Nadella: Much Critical Infrastructure Is 'Unpatchable' Amid AI Cyberattack Fears — TechCrunch AI · 2026-11-19
- WireWatch measures security of proprietary network encryption across the Android ecosystem — jedisct1 · 2026-11-19