CrowdStrike researchers have developed an automated system for classifying zero-day malware, addressing challenges in traditional methods that rely on manual analysis and malware family naming. By leveraging machine intelligence and behavioral data, the system uses supervised and unsupervised clustering techniques to categorize malware into pre-defined Threat Type groups, such as Adware, Backdoor, and Ransomware, among others. This approach enables rapid threat detection and enhances protection against unknown malware without the need for human intervention or consensus from third-party sources. The integration of a multi-class classifier further boosts confidence in threat identification, allowing for faster and more effective responses to new malware threats. The system's efficacy is demonstrated through a combination of centroid-based similarity assessments and K-means clustering to refine classification accuracy, all aimed at strengthening customer protection against cybersecurity threats.