A team of researchers has developed a new artificial intelligence model designed to analyze and verify potential cybersecurity threats to enterprise systems. The model, known as Cogent VR-1, is specifically trained for cybersecurity tasks and comes with tools to test and govern its performance. The system includes a benchmark to evaluate the effectiveness of cybersecurity agents and a runtime environment to manage their execution. This development aims to enhance the ability of organizations to identify and mitigate potential security risks.