Gray Swan AI Security Suite is an enterprise security platform designed to protect AI-powered applications from threats and vulnerabilities. The suite combines two main components: Cygnal for real-time AI input and output filtering, and Shade for automated testing. The platform provides AI red-teaming capabilities that pressure-test specific deployments using techniques discovered through research and Arena competitions. Vulnerability assessments are tailored to specific tools, data, and use cases. Gray Swan offers protection for AI agents and workflows, securing autonomous systems with tool access. The platform supports MCP and tool integration, protecting AI systems connected to databases, APIs, and files. It includes compliance and governance features to prepare for AI insurance and regulatory requirements. For AI labs and model developers, the platform provides private red-teaming services, custom research capabilities, and regulatory support aligned with emerging AI safety frameworks. The company conducts frontier AI research across evaluation, reliability, and control. Research areas include benchmarks like MMLU, WMDP, CyBench, HarmBench, and AgentHarm for evaluation. Reliability and control research includes GCG for jailbreaking testing, Circuit Breakers for adversarially robust alignment, RepE for representation engineering, Agent Red Teaming competitions, and Safety Pretraining interventions. Security researchers can participate in Arena competitions for discovering AI vulnerabilities, access a proving ground with weekly challenges, and utilize a research hub with curated tools and datasets.