Machine learning algorithms to identify and categorize cyber hazards have garnered attention recently. Malicious URLs, IP addresses, and files provide the greatest cybersecurity risks. This abstract describes how machine learning-based classification of dangerous URLs, IP addresses, and files identifies and mitigates threats. The article begins with hazardous URL categorization, which uses machine learning to identify if a URL is risky depending on its attributes. URL classification is challenging since there are many legitimate URLs. Researchers have developed many machine learning algorithms that assess URL domain names, length, and keywords to identify fraudulent URLs. Second, machine learning-based IP address classification is discussed. Classifying IP addresses helps identify and combat DDoS and botnet attacks. To discover rogue IPs, machine learning techniques like clustering, classification, and regression examine IP reputation, geolocation, and traffic patterns. File categorization—using machine learning to identify and categorize potentially hazardous files—concludes the study. If users open harmful files, viruses, Trojans, and other malware may infect their systems and steal personal data. Using file type, size, and behavior, supervised and unsupervised machine learning systems may identify dangerous files. This study discusses machine learning methods for detecting and banning risky websites and files. Machine learning algorithms can detect and mitigate cyber risks through analysis and categorization. These algorithms have false positives and negatives, overlook dangers, and require a lot of training data. Thus, machine learning must be employed with intrusion detection systems and firewalls to guard against cyberattacks.