Irregular

Irregular is an AI security lab working with the world’s leading frontier AI companies to understand and secure the most advanced AI systems. The company stress-tests frontier models and agents to uncover emerging security risks, unexpected behaviors, and vulnerabilities before they become real-world threats.

Irregular works closely with leading AI labs, including OpenAI, Anthropic, Google DeepMind, and Meta, helping them evaluate increasingly capable systems and build stronger safeguards. As AI agents become more autonomous and gain the ability to take actions in real-world environments, Irregular is building the security infrastructure needed to ensure they can be deployed safely and securely.