Microsoft fixes record 972 vulnerabilities as AI-enabled attacks become the new normal
Ars Technica AI · rss · 2026-09-09
Microsoft's September patch release fixes a record 972 vulnerabilities, 112 of them high critical severity — far above the record 570 patched just two months ago. Google and others have also seen record patch counts.
Context: two weeks ago OpenAI, Anthropic, AWS, Google, Microsoft and 100+ organizations signed an open letter warning of a narrowing window to patch bugs ahead of an expected tsunami of AI-enabled attacks that exploit them first.
Zero Day Initiative researcher Dustin Childs calls the spike the "new normal" and cautions AI-assisted attacks could eventually cause substantial damage.
More from Safety
- Split-LLM Training Privacy Failure: Zero-Gradient Patterns Leak Real Training Rows — Setloop · 2026-09-09
- After Altman's pacing-for-safety remark, Wildeford proposes concrete slowdown options for OpenAI — peterwildeford · 2026-09-09
- Peter Wildeford: Managing superintelligence will be more Cuban Missile Crisis than NPT treaty — peterwildeford · 2026-09-09
- NSA, FBI, CISA advisory accuses Chinese AI firms of industrial-scale distillation of US frontier models — Cklly2004 · 2026-09-09
- Fake BrightonSEO Talk Quote Shows AI Slop Now Feeding Back Into LLMs — lilyraynyc · 2026-09-09
- User reports Meta's new Muse bot bound his phone number to a stranger's account — BLUECOW009 · 2026-09-09