Topic

OWASP GenAI

News, analysis and practical guidance on OWASP GenAI.

2 articlesUpdated 30 Sep 2026

ResearchCloud & SaaS
Model Reasoning Trace Interception: Exfiltrating Proprietary Chain-of-Thought Data in Frontier AI
New research reveals side-channel and prompt-leakage vectors capable of intercepting hidden chain-of-thought reasoning tokens in frontier reasoning models.
CST Newsroom30 Sep 202610 min read
InsightsCloud & SaaS
The Lethal Trifecta in Enterprise AI Agents: Why Input Filters Fail Against Tool Hijacking
An architectural analysis of the Lethal Trifecta in autonomous enterprise AI agents: untrusted input, autonomous execution, and access to privileged external tools.
CST Newsroom30 Sep 202611 min read