Verno Labs

Verno Labs is a cybersecurity company focused on offensive security testing and runtime protection for AI agents. The platform is designed for enterprises deploying autonomous AI systems at scale, covering text, voice, and multimodal interactions. The platform consists of three core components: - Verno Red: Deploys custom offensive large language models (LLMs) to run autonomous attack campaigns against AI agents. These campaigns plan, execute, and adapt to discover vulnerabilities before attackers can exploit them. - Verno Guard: Provides real-time runtime detection of attacks in production environments, monitoring across agents and tool calls. - Verno Fix: Generates automated remediation recommendations and tracks findings from discovery through to closure. Underlying the platform is an intelligence layer that continuously learns from adversarial behavior across red teaming, runtime monitoring, and production telemetry. Verno Labs trains its offensive LLMs on proprietary adversarial datasets and maintains a proprietary attack taxonomy that maps AI-specific threats, exploit chains, and adversarial behaviors. The platform supports flexible deployment options, including self-hosted containerized instances within customer data centers or cloud environments (AWS, GCP, Azure), as well as a fully managed SaaS offering. An AI security agent interface allows users to query security findings and investigate vulnerabilities using natural language. The team has backgrounds from companies including Microsoft, AWS, and Expedia Group. Verno Labs is attending Black Hat USA 2026, indicating active engagement with the enterprise security community.