Vulnerabilities

Bad Vibes: Security Flaws Plague Popular AI Coding Agents

Tenzai study finds 72 vulnerabilities in code from AI agents including Cursor, Claude Code, and Codex. Business logic flaws dominate despite SQL/XSS prevention.

Read more →

Agentic Browsers Lack Isolation, Resurface Decades-Old Web Vulnerabilities

Trail of Bits research shows agentic browsers resurrect XSS and CSRF vulnerabilities through inadequate trust zone isolation between AI agents and web content.

Read more →

The Agent Security Paradox: Trusted Commands in Cursor Become Attack Vectors

Pillar Security reveals Agent Security Paradox where trusted Cursor commands become attack vectors through malicious repository content exploitation.

Read more →

Zero-Day Vulnerabilities Explained: Detection and Defense

Learn what zero-day vulnerabilities are, how attackers exploit them, and effective detection and defense strategies for your organization.

Read more →