When the Bee Stings: CyberCom’s AI Vulnerability

CyberCom, a cloud-based security company, uses a machine learning model (an AI system trained on data to recognize patterns) in its products to detect malware (malicious software). Security researchers found that this model fails to detect mutated malware (altered versions of malicious code), exposing a key weakness: AI tools that rely only on historical data can't effectively handle new or changed threats.