The Lasso AI Security Platform is an end-to-end cybersecurity solution designed to discover, monitor, and protect generative AI interactions in enterprise environments. It secures large language models (LLMs), AI agents, copilots, and retrieval-augmented generation (RAG) workflows by analyzing the semantic intent of user requests and agent actions. The platform enforces security policies and provides inline runtime guardrails to protect sensitive data between internal applications and users across various AI deployment environments.
Its technical capabilities include real-time threat detection for prompt injections, data leakage, anomalies, and unauthorized tool usage, with a reported classification latency of sub-50ms. The system offers automated adversarial testing (red teaming), agent and tool discovery, context-aware access controls, and data loss prevention (DLP). It also generates comprehensive audit trails to maintain visibility over AI usage, including shadow AI, and to support compliance reporting.
