
AI Agent Finds 21 FFmpeg Zero-Days Including Unauthenticated RCE
Depthfirst’s autonomous AI security agent spent $1,000 to find 21 zero-days in FFmpeg, including an unauthenticated RCE triggered by a

Depthfirst’s autonomous AI security agent spent $1,000 to find 21 zero-days in FFmpeg, including an unauthenticated RCE triggered by a

Anthropic has deployed six engineers inside NSA to operate Mythos, an AI reported capable of zero-day exploitation across major operating

Researcher Taylor Hornby used Claude Opus 4.8 to uncover a four-year-old Zcash Orchard flaw that could have enabled undetectable counterfeit

Volexity found Chinese APT VerdantBamboo used new PLENET and AGENTPSD malware to maintain 18 months of undetected Microsoft 365 access

JFrog researchers discovered IronWorm, a Rust-based infostealer with an eBPF rootkit, injected into 36 npm packages to steal AI API

AppEsteem found a Monero cryptominer bundled inside Hola Browser’s Windows installer, hidden as a Windows service and excluded from Windows

CISA added CVE-2026-45247 to its KEV catalog, confirming active exploitation of a CVSS 9.8 Magento RCE flaw that threatens 150,000

CVE-2026-8181 in Burst Statistics for WordPress is under mass exploitation, with Wordfence blocking 7,400 daily attempts against over 200,000 affected

Cisco confirmed public PoC code for CVE-2026-20230, a Unified CM SSRF enabling unauthenticated file writes and potential root access on

Team Xint Code used an AI tool to find CVE-2026-23479, a two-year-old Redis RCE posing high risk in cloud environments
Subscribe to the Daily Security Review Email Briefing to stay informed on the latest threats, trends, and technology, along with insightful columns from industry experts.