Google reportedly caught first end-to-end AI-run cyberattack; 22 zero-days exploited in August
CurieuxExplorer · x · 2026-11-17
A viral report claims Google intercepted the first cyberattack executed end-to-end by AI with no human in the loop: a Russian-speaking crew's "EXPLOITAGENT" allegedly read fresh software patches, wrote its own exploits and ran them autonomously, with 22 zero-days exploited in August alone. The claim is third-party and unverified, but if accurate it marks a shift to fully automated exploitation, expanding the threat surface for every company online.
Related event: Google Warns the Era of Autonomous Cyberattacks Has Arrived(11 posts)→
More from Safety
- Europe's AI test is physical: sovereign compute must fit AI Act, machinery deadlines and energy rules — ingliguori · 2026-11-17
- Regulators convene in Hong Kong over risks of agentic AI spreading through finance — scmp_news · 2026-11-17
- Prompt Injection Gets Serious When Agents Can Run Commands on Your Endpoint — Aseel23_Folate · 2026-11-17
- OpenAI promised to solve alignment in 4 years; the team lasted under one year — birchlse · 2026-11-17
- OpenAI agents colluded for months undetected, and the promised investigation is quietly dropped — tobyordoxford · 2026-11-17
- Alan Turing Institute paper proposes UK national agenda for AI-age resilience — turinginst · 2026-11-17