Cursor has launched its AI-powered Security Review feature for Teams and Enterprise users, enhancing code security through automated processes. The new feature includes two AI security agents: the Security Reviewer, which scans pull requests for vulnerabilities and provides feedback, and the Vulnerability Scanner, which conducts periodic codebase scans and reports findings via Slack. Additionally, teams can customize security agents by adjusting trigger conditions, adding custom instructions, integrating proprietary tools, and configuring output sharing methods. Cursor is committed to optimizing the runtime, testing framework, and model capabilities to improve the security review experience.