Mar 15, 2025·8 min·Featured
Adversarial Attacks on Vision Language Models
Exploring typographic and visual prompt injection techniques that bypass VLM safety guardrails through carefully crafted image perturbations.
AI SecurityResearch
/07 — Dispatches
BLOG
Notes on offensive security, AI research, intelligence engineering, and the tools we build. Sporadically updated.
Feb 28, 2025·6 min
Subdomain Takeover at Scale: Automating Discovery
How we built SecRecon's subdomain enumeration engine to identify dangling DNS records across Fortune 500 attack surfaces.
Offensive SecurityTooling
Feb 10, 2025·10 min
Building RAG Pipelines for Threat Intelligence
Using retrieval-augmented generation to correlate CVE databases, dark web feeds, and internal telemetry into actionable intel.
AIThreat Intel
Jan 20, 2025·12 min·Featured
Cognitive Purple Teaming: Autonomous Attack Simulation
VANGUARD's approach to generating attack chains, deploying KQL detection rules, and measuring SOC response in real-time.
Purple TeamAutomation
Jan 5, 2025·9 min
Extracting Behavioral Signals from Open Source Intelligence
A framework for processing SOCMINT data streams to identify behavioral patterns, influence operations, and potential insider threats.
OSINTIntelligence
5 ENTRIES · MORE COMING SOON