LAB42AI

> AI agents and agent infrastructure security

SERVICES & CAPABILITIES
โš”๏ธ
RED TEAM
AI Agent Attack Simulation
๐Ÿค–
DEVELOP
Build, Deploy & Manage AI Agents
๐Ÿ“„
RESEARCH
AI Agent Vulnerability Research
๐Ÿ› ๏ธ
TOOLING
Open Source Offensive AI Agent Exploits
VENTURES
๐Ÿข Soteria
Fully managed AI agent infrastructure for businesses that won't put sensitive client data on shared servers or someone else's model. Private hardware. No cloud lock-in. No DevOps.
[STATUS: LIVE]
[soteriainfra.com]
ARSENAL
โš”๏ธ ZeroShot
A red teaming tool for AI systems. Probe, stress-test, and break LLMs before attackers do.
[STATUS: ONLINE]
๐Ÿค– Minion
A security-first AI agent built with safety guardrails from the ground up.
[STATUS: ONLINE]
RESEARCH & TALKS
TALK
Talk at Louisville AI Week
Louisville AI Week ยท February 2025
TALK
MLOps Virtual Conference ยท November 2025
PAPER
Venue / Journal ยท 2025
ABOUT US

We are a security research lab focused on AI agents and the infrastructure behind them.

We break agent systems, we find the failure points, then we build open-source toolkits so the security community can do the same. Red team first, always.