Microsoft's record Patch Tuesday drew seven CVE counts. Security vendors use AI to rank them, and Ivanti's Claude skill invented details reviewers caught.
CrowdStrike's Falcon Guardian found 18,000 AI agents on one Fortune 500 network — the company had approved just 300, exposing how little security teams see.
Infostealers harvested Claude session cookies and replayed them past 2FA, and some hijacked accounts held standing access into corporate Gmail and Drive.
Prompt injection is OWASP's No. 1 LLM risk, but ranks No. 12 in 6,639 real incidents — an independent analysis finds no statistically detectable agreement.
Anthropic gave three Claude agents conflicting orders on one server. They sabotaged each other, disguised malware, and hid it from the humans relying on them.
For decades Mastercard's fraud rules were built to stop bots. Now AI shopping agents are buying, forcing Mastercard to rewrite its core risk framework.
AI model verification has relied on ClamAV scan badges and a repository tag no one has to prove. Cisco's new tool grounds lineage in weight-level fingerprinting instead.
Rajat Taneja explains why Visa open-sourced the AI harness that found critical bugs in its payment network, and the new metric it built to verify fixes work.
