General Analysis
General Analysis builds the security layer for agentic systems, providing automated red-teaming, runtime guardrails, and asset management for enterprise AI.
At a Glance
- Enterprise technology companies
- Financial institutions
- AI research and development labs
AI Tools by General Analysis
(1)
General Analysis
AI Red Teaming for Agentic Systems
Discussions
No discussions yet
Be the first to start a discussion about General Analysis
Latest News
Supabase MCP can leak your entire SQL database research
50+ customer service agents offer $10,000,000+ in fabricated benefits research
General Analysis releases the GA Guard series of safety classifiers
Claude Jailbroken to Mint Unlimited Stripe Coupons research
Products & Services
A context-aware red teaming platform that maps tool graphs and generates multi-step exploits to identify system-level risks.
A family of adversarially trained safety classifiers and guardrails for enterprise AI deployments.
Tools for full visibility and scanning across AI infrastructure, models, and knowledge assets.
A free runtime firewall designed to protect against vulnerabilities in the Model Context Protocol (MCP).
Market Position
General Analysis positions itself as the security layer for the 'agentic era', focusing on system-level risks and multi-step exploits that traditional security tools miss.
Leadership
Founders
Rex Liu
Co-founder. Previously at Caltech and CMU. Focuses on AI security and safety research.
Rez Havaei
Co-founder. Background in AI security and safety from Carnegie Mellon University (CMU).
Alan Wu
Co-founder. Previously at Caltech and Jump Trading. Background in quantitative trading and computer science.
Executive Team
Rex Liu
Co-founder
Caltech, CMU research background.
Rez Havaei
Co-founder
Carnegie Mellon University (CMU) research background.
Founding Story
General Analysis was founded by a team from Caltech, CMU, and quant trading who observed that organizations were racing to deploy AI without sufficient safety infrastructure. They realized that autonomous agents invert the traditional security model, requiring a specialized security layer.
Business Model
Revenue Model
Enterprise SaaS/Security Platform. Likely includes subscription fees for advanced red-teaming and defense tools, plus usage-based licensing for enterprise guardrails.
Pricing Tiers
Runtime firewall for MCP vulnerabilities.
Full suite including automated red-teaming, asset management, and advanced guardrails.
Target Markets
- Enterprise technology companies
- Financial institutions
- AI research and development labs
- Automated AI red-teaming
- CI/CD security evaluation
- Runtime defense for AI agents
- AI asset vulnerability scanning
- Indeed
- Together AI
- NVIDIA
- Jane Street