One week in April 2026 delivered a CVSS 9.8 AI-agent flaw, an unauthenticated SGLang RCE, MCP memory poisoning, and a 766-host credential harvest. Here's the full analysis and your patching order.
RSAC 2026's dominant theme was agentic AI security — AI agents are now the target. Cisco's DefenseClaw, the MCP server problem, and a five-step playbook for securing autonomous agents.
A comprehensive analysis of the top 4 AI agent security threats in 2026: tool-use exploitation, supply chain attacks, cross-agent contamination, and model extraction — with actionable defense strategies.